Sei sulla pagina 1di 6
Journal of Cleaner Production 73 (2014) 269 e 274 Contents lists available at ScienceDirect Journal of

Contents lists available at ScienceDirect

Journal of Cleaner Production

journal homepage: www.elsevier.com/locate/jclepro The environmental effects of seasonal food purchase: a

The environmental effects of seasonal food purchase: a raspberry case study

Chris Foster a , * , Catarina Guében a , Mark Holmes b , Jeremy Wiltshire b , Sarah Wynn b

a EuGeos Limited, UK

b ADAS, UK

b , Sarah Wynn b a EuGeos Limited, UK b ADAS, UK article info Article history:

article info

Article history:

Received 28 March 2013 Received in revised form 6 December 2013 Accepted 19 December 2013 Available online 17 January 2014

Keywords:

Seasonal

Fruit

Seasonal food

Raspberry

LCA

abstract

The environmental effects of seasonal food supply have been explored through a Life Cycle Assessment (LCA) study of raspberries supplied to UK consumers at different times of year. Supply of raspberries at different times of the year draws on different production systems and locations. Despite that, the results of this LCA, based on data from individual producers, reveal relatively small differences in impacts for different times of supply, except in the case of the water footprint measures. LCIA results are very sensitive to fruit yield. So in this case, yield and agricultural practice appear stronger drivers of the environmental burden of food production than is time of supply. In such situations a strong focus on seasonality in sustainable food provisioning is unlikely to deliver large environmental bene ts. Using LCA to establish what bene ts might be available from a more general shift to seasonal food con- sumption, often advocated as more sustainable , will require a multi-product approach. Such an approach could take current food consumption patterns or environmental targets as its starting point. 2014 Elsevier Ltd. All rights reserved.

1. Introduction

As urbanisation progressed in the second half of the 20th Cen- tury and the agricultural workforce shrank, so Western European citizens disengaged from food production, losing their connection with its seasonal patterns. From the 1960 s onwards, seasonal variation in the availability of certain foods reduced, leading to the commonly-described position of all-year-round availability for many foods. Recently, interest in seasonal foods has been resur- gent; Dibb et al. (2006) state that two-thirds of people in the UK are now taking steps to buy seasonally . This trend has various drivers but e as Dibb et al. s title suggests e some see implications for the environment in it. In line with this, advice on sustainable diet often advocates consumption of seasonal food. Seeking additional evidence relevant to such recommendations, the UK s Department of the Environment, Food and Rural Affairs commissioned a research project exploring the environmental implications of sea- sonal food purchasing (DEFRA research project FO0412). This paper reports some ndings of the project, focussing on environmental implications of seasonal food supply explored through a raspberry LCA case study.

* Corresponding author. E-mail addresses: chrisf@eugeos.co.uk , c.foster@manchester.ac.uk , chris.foster@ mbs.ac.uk (C. Foster).

0959-6526/$ e see front matter 2014 Elsevier Ltd. All rights reserved.

1.1. Seasonalfood

A review of literature and consumer research demonstrated that clearly identifying seasonal food is in fact quite dif cult. Few commentators take the trouble to de ne the term seasonal , while consumer research found that UK consumers have only a vague de nition of seasonal food . In essence very different de nitions and perceptions of what is seasonal are applied by different parties ( Brooks and Foster, 2011 ). To inform the project noted above, two working de nitions of seasonal food were used e one derived from discussions with industry and policy makers and one informed by consumer research reported in ADAS et al. (2012) . The rst was a production-oriented or global de nition: food that is outdoor grown or produced during the natural growing/production period for the country or region where it is produced. It need not neces- sarily be consumed locally to where it is grown. 1 The second was a consumer-oriented, more local de nition: food that is produced and consumed in the same climatic zone, e.g. UK, without high energy use for climate modi cation such as heated glasshouses or high energy use cold storage. Inevitably, these de nitions are themselves open to interpretation. The LCA research on which this article draws covered a number of food items meeting one or the other (or both) of these de nitions.

1 This was originally suggested in Defra s project speci cation.

270

C. Foster et al. / Journal of Cleaner Production 73 (2014) 269e 274

1.2. Food, seasonality and the environment

The timing of agricultural activities in any one place can change the effects of those activities on the wider environment, even if the activities remain the same. Thus changing the timing of pesticide applications can result in increased or reduced effects on non-pest susceptible species, simply because these will be present in different numbers, variety and development stage; moving nitro- gen fertiliser applications to times of higher rainfall will likely lead to higher leaching rates; the presence of crop canopy in times of higher rainfall might mitigate soil erosion. So if a crop s planting-to- harvest cycle is moved earlier or later in the year and fertiliser and/ or pesticide applications moved in step, then the total effects on the environment can change. As food production for supply in a certain place is shifted further away in time from the natural , or normal time of production there, so one or both of two additional changes occurs: either the nature of the producing activity changes (e.g. through the use of protected growing environments) or the place of production changes (change can include division, as in the case reported below). Furthermore, preservation and storage allow the time of production and the time of supply to be separated, introducing further exibility into the supply system. Finally of course, con- sumers also have access to preservation and storage, so can sepa- rate the time of supply from the moment of consumption. Each of these adjustments changes the interaction between the food system and the natural environment surrounding it: different production systems for the same basic foodstuff have different yields and require different inputs, almost all preservation tech- niques require energy inputs, as does cold storage. The fact that these adjustments can be made at different points in a generic food production-consumption system is a strong indicator that life cycle assessment will be an effective tool to explore their environmental implications.

1.3. LCA and seasonality

While the need to consider the whole food production- consumption system favours the application of LCA in this context, careis needed because of the way temporality is handled within LCA. For agricultural products the product system is normally de ned as a full annual (sometimes multi-annual) cycle and the life cycle in- ventory (LCI) integrates emissions occurring throughout that cycle, dividing them equally among the total harvest. Thus few life cycle analyses of food products or systems explicitly explore seasonality. When LCA studies refer to seasonality , the term is associated e implicitly or explicitly e with a crop s natural growing season and its cropping period , thus with its availability for fresh consumption. Some LCAs examine seasonal variation in impacts more explicitly, for example Williams et al. (2009) . This and other studies ( Blanke 2007; Blanke and Burdick, 2005, 2007, Hospido et al., 2009; Jones, 2006; Milà i Canals et al., 2007; Saunders et al., 2006 ) consider seasonality in the context of the supply of fresh produce to con- sumers in Northern Europe all year round e and thus closely con- nected to the local vs. global or food miles debate. Examination of this literature highlights some issues that require attention if seasonal effects on the environmental impacts of food supply are to be separated from other factors that may differ between supply systems but are not directly linked to seasonality. For example, a review by Evans (2014) found the range of speci c energy use (energy use per volume) in cold stores in the UK alone to be very wide for each temperature regime studied e with an eight- fold difference between the most and least ef cient. Blanke and Burdick (2005), Milà i Canals et al. (2007), Saunders et al. (2006), Sim et al. (2007) and Williams et al. (2009) all report LCAs of

al. (2007) and Williams et al. (2009) all report LCAs of Fig. 1. UK raspberry supply

Fig. 1. UK raspberry supply 2007. Sources: Defra horticultural statistics, UK Trade statistics.

apples; it is clear from these that the impacts associated with the different stages of the apple life cycle are of similar order of magnitude, a situation that may then reasonably be anticipated for other top-fruit. Differences in post-harvest technology, arising coincidentally, may therefore outweigh differences driven by the season of production. In cases where supply at a particular time of year requires storage, the scale of product loss or degradation during storage must be accounted for. The degree of geographical resolution embedded in impact assessment (LCIA) methods (few of which exist in regionalised form) may also limit the extent to which LCA can inform about the environmental effects of seasonal varia- tion in food supply when that variation involves production in different places.

2. Methods

2.1. Scope

The aim of the research was to explore the environmental im- plications of upstream changes that arise as supply of particular foodstuffs progresses through the year. Therefore a selection of individual foods was studied, rather than a sequence of baskets . Here the raspberry case study is reported to illustrate how envi- ronmental impacts vary across the year for one food consumed in the UK. Clearly at a certain time of year raspberries are in-season in the UK, at other times they are not. The project considered only the effect of changing the times of production and supply in the system as far as delivery to the food retailer. In effect, we equate (reecting mainstream economics and consumer data) consump- tion with purchase, and purchase with supply to the retailer. This embodies a simpli cation: it is possible that consumers store foods for extended periods after purchasing them. The environmental implications of this, if it occurs, were not considered in the LCA; it would make food consumption less seasonal than statistics would lead us to believe it is. Some of the volume captured by this data is supplied to commercial buyers (the foodservice sector or institutions) rather than nal consumers, of course. This is still purchase, however, and there seems to be no reason to exclude it. Fig. 1 shows how UK supply of raspberries changes through the year in volume and by source (data compiled from UK production 2 and import 3 statistics with quantity, in tonnes, as the y-axis which

2 Department of Environment, Food and Rural Affairs Horticultural Statistics:

3 HM Revenue & Customs Trade Statistics: https://www.uktradeinfo.com/Pages/ Home.aspx .

C. Foster et al. / Journal of Cleaner Production 73 (2014) 269e 274

271

is not shown) There is scarcely competition between local pro- duction e certainly seasonal according to the consumer-oriented de nition above e and imports, which are seasonal only ac- cording to the global de nition; rather imports complement local produce in an overall supply pattern. To gain some insight into the environmental implications of this supply pattern an LCA of rasp- berries was conducted. This covered 3 functional units:

A. 1 kg raspberries delivered fresh to a supermarket s retail dis- tribution centre (RDC) in May

B. 1 kg raspberries delivered fresh to a RDC in July

C. 1 kg raspberries delivered frozen to a RDC in November

Given the data underlying Fig. 1, an appropriate product system for A involves production in Southern Spain. In a production sys- tem common there, raspberries are grown on an annual basis in elds that are covered for the whole season with Spanish tun- nels. The ground is prepared each year and beds then formed. The planting material (canes) is produced in the UK or Netherlands and transported to the producing site in chilled lorries. The canes are kept for 3 e 4 weeks in a cold store prior to planting, then planted directly through plastic into the pre-prepared beds. Fer- tilisers are applied through drip irrigation: nitrogen as ammonium nitrate or potassium nitrate. Water for irrigation is abstracted from an aquifer. Harvesting occurs by hand with the fruit then trans- ported directly to the packhouse; average yield is 8 tonnes marketable fruit per ha. (FAOSTAT 2009) In the packhouse the fruit is graded and cleaned, before being placed into punnets with a plastic lm lid. Punnets are cooled in a cold store prior to export in refrigerated trucks which travel 2,500 km from Spain to the UK. For B and C production in the UK is clearly appropriate. a typical production system is located in the east of England, on canes grown for seven years in elds covered with polytunnels during fruiting. In this case in the rst year the ground is prepared, beds formed and soil sterilised (e.g. with chloropicrin). The canes are produced on a separate farm, cold stored prior to planting, then planted directly through plastic into the pre-prepared beds. On an annual basis (for seven croppings) fertilisers are applied through drip irrigation including nitrogen fertiliser as ammonium nitrate or potassium nitrate. Irrigation uses water from an aquifer. Fruit is hand-harvested then transported directly to the packhouse. Yield averages 12t marketable fruit per hectare per year for the seven cropping years. In the packhouse the fruit is graded and cleaned, before being placed into punnets with a plastic lm lid. Punnets are cooled in a cold store prior to distribution in refrigerated trucks. At the end of the season the soil between the beds is pulverised to reduce compaction. At the end of the seven years the crop is grubbed out, the ground sub-soiled and the plastic rolled up and recycled. These are commonly-encountered production techniques pro- ducing for the UK supply pattern, and individual producers employing them provided complete or partial operating data for use in the project (see next section). In the case of C, Fig. 1 indicates that there is little e although not zero e supply from primary producers in November. Supply of frozen raspberries originally produced in the UK production system described for B was included to investigate the signi cance of adding storage alone to one of the production systems already studies. An alternative mechanism for supply of raspberries to the UK market in November would involve import by air of fresh fruit produced overseas, probably North America on the basis of Fig. 1. We had no access to data characterising raspberry production in North America and so this possible fourth case was not included. Frozen and fresh raspberries may indeed not be considered as substitutable products by purchasers.

2.2. Boundaries

The product systems incorporated production of fertilisers, canes, packaging, fuels and all other inputs. Production of material for polytunnels was included, but other capital equipment was excluded so that the calculations of global warming potential were compliant with PAS 2050:2008, which was the most recent version of that standard at the time the work was carried out. For the production system in B & C, an additional year of operation without any crop production was included as an allowance for cane pro- duction, for which direct data were unavailable.

2.3. Data

Primary data from individual operating locations were used to characterise agricultural operations, with expert consultation used to ll data gaps. Background data to represent production of inputs such as fertilisers, fuel and polyethylene lm were taken from the ecoinvent database v2 ( www.ecoinvent.ch ). Data characterising outdoor production of raspberry canes in the UK were developed by the research team, based on the data provided for raspberry fruit production outdoors in the UK and advice from a horticultural expert on likely per ha yields. Methane and nitrous oxide emis- sions, where they arose, were calculated using the method set out in the IPCC s (2006) Guidelines for National Greenhouse Gas In- ventories. This aligned the method with PAS 2050:2008. The water footprint of a primary crop is calculated as the ratio of volume of water consumed for crop production to the crop yield. The Food and Agriculture Organization of the United Nations CROPWAT tool ( FAO, 2010a ) was used to estimate volumes of rain and irrigation water evaporated in the eld during production: the green and blue water footprint. As input into this model, climate data was taken from the CLIMWAT database ( FAO, 2010b ), re ned with local data wherever possible. To calculate grey water footprint and to enable calculation of eutrophication potentials, nitrate leached to water was estimated using the ADAS Nitcat ( Lord, 1992 and subsequent revisions) eld- scale nitrate leaching model, with data speci c for each location. Where the model was deemed inappropriate for the growing conditions or crops, estimates of leached nitrate were obtained from the literature.

A single set of data, from one of the operating locations, was

used to characterise all packhouse operations and packaging; a loss rate of 1% in the packhouse was assumed.

A dataset for chilling and short-term cold storage of soft fruit

was developed to characterise packhouse operations and storage prior to transhipment either to the RDC or to a long-term cold- storage location. This was based on data for annual operation provided by one business participating in the study. To create an appropriate dataset for an equivalent operation in Spain, the source of electricity used was changed from the UK grid to the Spanish one; energy consumption was assumed to be the same in both locations e i.e. any additional cooling energy required in the warmer climate of Spain was ignored. A dataset characterising frozen storage was also developed, encompassing energy use and refrigerant loss and accounting for the burdens of providing a unit volume of cold storage capacity for a unit time; energy use in cold stores was taken from a review of UK facilities ( Evans, n.d. ); a mixture of gases (the constituents of R404A in the appropriate proportions and NH 3 ) was used to represent emissions to air, re ecting the most widely-used refrigerants, while actual data for refrigerant use (assumed equal to losses over a full year) were supplied by the raspberry packer. In the survey of energy use in cold stores, Evans (2014) found the range of speci c energy use within each temperature regime to be much greater than differences be- tween the different temperature regimes (i.e. frozen and

272

C. Foster et al. / Journal of Cleaner Production 73 (2014) 269e 274

chilled ). Because most LCA results for food products relate to a unit mass of product rather than volume (the factor to which cold store energy consumption is typically normalised for reporting purposes), gures for mass of product per unit of storage volume were taken from Brunel University (2009) to calculate the cold storage volume needed for 1 kg raspberries. Finally, a dataset for the unit operation of road transport in refrigerated trucks was developed using values for fuel consumption, vehicle utilisation and refrigerant loss given by Brunel University (2009) ; emissions per unit of fuel used were taken from ecoinvent v2.1 ( Spielmann et al., 2007).

2.4. Impact assessment

Impact assessment was conducted for environmental categories deemed relevant to the project. Category and impact assessment method selection were also guided by a requirement for compati- bility with results from previous LCAs of UK agricultural com- modities, particularly those produced by Williams et al. (2009) . Therefore, CML midpoint methods ( Guinée, 2002 ) were used for the categories global warming, eutrophication, acidi cation and photochemical oxidation. In addition agricultural land occupancy was reported in units of m 2 yr. (the result obtained by simply adding all land occupation of all classes in the inventory), as were the unweighted water footprint (to the method described by Hoekstra and Chapagain, 2008 ); weighted water footprint (using a method very similar to that outlined by Ridoutt and P ster (2010) whereby a water stress characterisation factor for producing loca- tions was introduced through the use of a water stress index (WSI, P ster et al., 2009 ) to weight the water use according to the de- gree of water stress at the place of use) and environmental impact quotient (EIQ, Kovach et al., 1992 ) of pesticides used. This last method is based on average effects, and is neither geographically- nor temporally-sensitive. Regionalised LCIA methods are available for eutrophication and acidi cation, but were not used here because other factors, noted above, drove LCIA method selection.

3. Results

The impact assessment results are shown in Table 1. Sensitivity analysis was undertaken to assess the effect on the LCIA results of raspberry cane yield (no. Per ha) for case A and of cold-store operating parameters for case C. Selected results are shown in Table 2 (below). Results for impact categories driven strongly by horticultural process parameters (for example, land occupation, global warming) are, as expected, highly sensitive to product yield. Categories to which transport process emissions contribute more (notably acidi cation) are of course less sensitive to this factor. Sensitivity to other parameters, such as cold-store

ef ciency, was found to be weaker than to those included in Table 2 . For cold storage, small variations in energy consumption per unit volume are less signi cant in uences on overall LCA re- sults than the nature of the refrigerant used and assumed loss rates for the more environmentally-signi cant refrigerants.

4. Discussion

The results in Table 1 represent the progression of the envi- ronmental impacts associated with supply of raspberries through the year. Because food at one time of the year is not, for many consumers, substitutable for food at another, while it is reasonable to re ect on the changes in LCIA results between one case and another, the three cases cannot really be compared as alternatives. The differences between the LCIA results obtained for these particular cases of raspberry supply at three points in the year are relatively small, except for the water footprint measures. The fact that both canes and fruit are subject to long-distance refrigerated transport is a signi cant factor behind the higher acidi cation and abiotic depletion values obtained for raspberries delivered fresh in May (A); for example product transport accounts for 35% of the acidi cation potential for A but 18% of the acidi cation potential for B. The close similarity between the EIQ values obtained partly re- sults from the use of expert consultation to ll data gaps; results obtained for strawberries within the same project (for which the pesticide use data was of higher quality) suggest that differences in soil sterilant application rates and frequencies can have a signi - cant in uence on this indicator. It is notable that the weighted water footprint measure is the one LCIA method applied here which is sensitive to location and is the one for which the differ- ences between the three cases is largest. For impact categories other than water, likely (e.g. year-to-year) variations in fruit yield and cane yield could give rise to variations in the results obtained for one particular case greater than the differences between the different cases shown in Table 1. N 2 O emissions from horticulture contribute a large proportion of the GWP: 75% for A and more than 90% for B and C. However, in this project the calculation of N 2 O emissions from soil following the incorporation of crop residues both in the UK or overseas was highly problematic. The IPCC 2006 method using the tier 1 approach is complex and uses many default values for speci c crops or crop groups. A large number of crops are not represented in the IPCC method, therefore default data relevant to several of the products considered in this project e including raspberries e were not available. Data for another crop product were used as a proxy, but this introduces a further element of uncertainty. The production of polyethylene contributes some 25% of the abiotic depletion potential in A. Tunnels account for the majority of the polyethylene in this case. While there is some uncertainty about

Table 1 LCIA results, raspberries in the UK at different times of year.

Impact category

Product system

A. Raspberries, fresh at UK RDC in May

B. Raspberries fresh at UK RDC in July

C. Raspberries frozen at UK RDC in November

GWP100 (kg CO 2 eq) Water footprint (WF) (m 3 Virtual water) Weighted WF (m 3 Virtual water) Agricultural fruit-growing land occupation (m 2 yr) Pesticide hazard indicator E.I.Q. Abiotic depletion (kg antimony eq.) Photochemical oxidation e high NO x (kg ethylene eq.) Acidi cation e (kg SO 2 eq.) Eutrophication (kg PO 4 eq.)

7.3

7.4

7.7

2.7

1.3

1.3

2.7

0.09

0.09

1.1

1.2

1.2

0.3

0.3

0.3

0.01

0.004

0.006

0.0004

0.0001

0.0002

0.01

0.003

0.004

0.005

0.004

0.004

C. Foster et al. / Journal of Cleaner Production 73 (2014) 269e 274

Table 2 Sensitivity analysis results.

273

Impact category

Product system

A. Raspberries, fresh at UK RDC in May (base)

A1. Raspberries, fresh at UK RDC in May, fruit yield þ 10%

A2. Raspberries fresh at UK RDC in May, planting material (canes) yield þ 30%

B. Raspberries fresh at UK RDC in July

B1. Raspberries fresh at UK RDC in July, fruit yield þ 10%

C. Raspberries frozen at UK RDC in November

C1. Raspberries frozen at UK RDC in November, maximal cold-store occupancy

GWP100 (kg CO 2 eq) Water footprint (WF) (m 3 Virtual water) Weighted WF (m 3 Virtual water) Agricultural fruit- growing land occupation (m 2 .yr) Pesticide hazard indicator E.I.Q. Abiotic depletion (kg antimony eq) Photochemical oxidation e high NO x (kg ethylene eq.) Acidi cation e (kg SO 2 eq.) Eutrophication (kg PO 4 eq.)

7.3

5.7

6.0

7.4

6.7

7.7

7.5

2.7

Not calculated

Not calculated

1.3

Not calculated

1.3

Not calculated

2.7

Not calculated

Not calculated

0.09

Not calculated

0.09

Not calculated

1.1

1.0

1.0

1.2

1.1

1.2

1.2

0.3

0.3

0.3

0.3

0.3

0.3

0.3

0.01

0.01

0.01

0.004

0.004

0.006

0.005

0.0004

0.0004

0.0004

0.0001

0.0001

0.0002

0.0002

0.01

0.008

0.009

0.003

0.003

0.004

0.003

0.005

0.004

0.005

0.004

0.003

0.004

0.003

the fate and longevity of the material used for these tunnels in practice, extending the material s life and recycling it when it is no longer useable are clearly desirable. In case C (frozen raspberries supplied in November) no allow- ance was made for loss or spoilage during cold storage. Such losses increase the impacts associated with supplied product but no relevant data for loss rates were available when the work was conducted. Recent work by WRAP ( Terry et al., 2011 ) provides an estimate of 2 e 3% losses of fresh raspberries in packing and in retail stores, but provides no estimate for losses of packed fruit consigned to frozen storage. The loss rates found for packing are similar to those used in this study.

5. Conclusions

An LCA has been completed of a soft fruit supplied in the UK at three different times of the year. Here we draw some tentative conclusions based on the results of this LCA. The impact assessment results obtained show relatively small shifts as the time of supply progresses through the year, perhaps surprising in light of the operational differences between the supply systems. The in uence on environmental impact of the place of production shows through strongly in the weighted water footprint. This impact assessment method has, of course, location- sensitivity built into it; it may be that if regionalised methods had been used for other categories (notably eutrophication), the in u- ence of place would have shown in those too. Comparing the results obtained for the different cases with the results of the sensitivity analysis suggests that, for raspberries at least, yield and agricultural practice are stronger drivers of the environmental burden of food production than is time of supply. In such situations, it seems that a strong focus on seasonality in sustainable food provisioning is unlikely e by itself e to deliver large environmental bene ts. That conclusion must be recognised as provisional because of the nature of the LCAs undertaken here e based on particular examples of production rather than statistically-representative datasets for production at certain times of the year, employing modelled data for emissions and LCIA methods without regional sensitivity to explore

the effects of emissions which arise in different places at different times of the year. Such reservations about LCA method aside, the literature review and study design highlighted challenges facing any assessment of the environmental effects of a shift towards seasonal food consumption . If such a shift occurred at any signi cant scale, it would presumablyinvolve a complex adjustment of food purchasing and consumption patterns. The method used in this work precluded consideration of the environmental consequences of seasonally- changing consumption patterns. Understanding the environ- mental implications of such patterns remains a desirable aim. Future work should look beyond single food items. An important rst step in such work would be to draw on food purchasing statistics to establish how consumers food purchases actually vary through the year; month-by-month results from the UK s National Food Survey represent such a source (Of ce of National Statistics). In setting up scenarios that might be used as realistic alternatives to existing patterns that might be revealed in such statistics, account must be taken of differences in understandings of seasonal food, modern health norms (e.g. 5-a-day ) and 21st-century consumer expecta- tions. 4 Established or desirable environmental goals should also be considered in any such scenario-based LCA study; such a study might then try to build year-long consumption patterns that meet both modern health norms and environmental targets while incorporating seasonal variation.

References

4 For a discussion of the co-evolution of consumer expectations and food prod- ucts, see Foster et al. (2012) or Freidberg (2009) .

274

C. Foster et al. / Journal of Cleaner Production 73 (2014) 269e 274

Jones, A., October 2006. A Life Cycle Analysis of UK Supermarket-imported Green Beans from Kenya. International Institute for Environment and Development (IIED). Fresh Insights No. 4 . Kovach, J., Petzoldt, C., Degnil, J., 1992. A method to measure the environmental impact of pesticides. New. Yorks Food Life Sci. Bull. 139, 1 e8 . Lord, E.I., 1992. Modelling of nitrate leaching: nitrate sensitive areas. Aspects Appl. Biol. 30, 19 e 28 . Milà i Canals, L., Cowell, S.J., Sim, S., Basson, L., 2007. Comparing domestic versus imported apples: a focus on energy use. Env. Sci. Pollut. Res. 14 (5), 338 e 344 . P ster, S., Koehler, A., Hellweg, S., 2009. Assessing the environmental impacts of freshwater consumption in LCA. Environ. Sci. Technol. 43, 4098 e 4104 . Ridoutt, B.G., P ster, S., 2010. A revised approach to water footprinting to make transparent the impacts of consumption and production on global freshwater scarcity. Global Environ. Change e Hum. Policy Dimen. 20 (1), 113 e120 . Saunders, C., Barber, A., Taylor, G., July 2006. Food Miles e Comparative Energy/ Emissions: Performance of New Zealand s Agriculture Industry. Agribusiness & Economics Research Unit, Lincoln University, New Zealand, ISBN 0-909042-71- 3. Research Report 285 . Sim, S., Barry, M., Clift, R., Cowell, S.J., 2007. The relative importance of transport in determining an appropriate sustainability strategy for food sourcing. Int. J. Life Cycle Assess. 12 (6), 422 e431. Spielmann, M., Bauer, C., Dones, R., Tuchscmid, M., 2007. Transport Services. Ecoinvent report no. 14, Villigen and Uster. Terry, L.A., Mena, C., Williams, A., Jenney, N., Whitehead, P., 2011. Fruit and Vege- table Resource Maps. WRAP, Banbury. Final report of project RSC008 . Williams, A.G., Pell, E., Webb, J., Tribe, E., Evans, D., Moorhouse, E., Watkiss, P., 2009. Comparative Life Cycle Assessment of Food Commodities Procured for UK Consumption through a Diversity of Supply Chains. Defra, London. Final Report for Defra Project FO0103 .