Emergy Synthesis of Food Preparation and Diets in the“Green”Urban District Rosendal,in Uppsala,Sweden

Jacinda J.Maassen,Torbjrn Rydberg,Daniel Bergquist

Department of Urban and Rural Development,Swedish University of Agricultural Sciences,Almas All′e 8,750 07,Uppsala,Sweden

Keywords Urban development Sustainability Environmental support Food systems Emergy Urban diets

Abstract

1 Introduction

Global trends show that as nations urbanize and become wealthier, they converge to a “Western” diet (Ranganathan et al., 2016). A Western diet is defined as the intake of higher amounts in calories, protein, and animal products (ibid.). Therefore, in growing economies and urban areas, the quantity of food consumed per capita increases and the source of nutritional energy shifts from carbohydrates to animal products, fats and oils(Gerbens-Leenes et al.,2010). For example,in China an increase in the consumption of meat and pork,fat and eggs and a decrease in the consumption of grain are associated with urban and high-income populations(Guo et al., 2000; Ma et al., 2006; Kearney, 2010). This shift is problematic as research shows that meat consumption in particular is linked to the use of more environmental resources and higher greenhouse gas(GHG)emissions than other foods (Weber and Matthews, 2008; Scarborough et al.,2014; Tilman and Clark,2014). In addition,McLaughlin and Kinzelbach(2015)argue that there may not be enough resources to feed future populations on such diets since they are commonly associated with poor environmental management and production practices.

Various researchers show that food production and consumption in general have substantial effects on the environment and the use of resources. For example,Tilman and Clark(2014)show that urban and affluent populations consume higher quantities of total and empty calories as well as foods that have higher environmental impacts than their rural counterparts (Kearney, 2010; Deelstra and Girardet, 2000). Hence,urban diets require more environmental support and the use of more resources. With urbanization and affluence on the rise, these trends become increasingly important. Yet,urban food preferences and diets as well as agricultural production practices are most often neglected in urban planning and policy(Deelstra and Girardet,2000). Urban agriculture is sometimes proposed as a strategy to improve access to fresh foods, social cohesion and collective efficacy.However, the implementation of local food systems has been limited because it often falls through the cracks in comprehensive land-use planning and sustainability plans (Wekerle, 2004; Thibert 2012). Furthermore, few studies address issues of sustainability and resource efficiency related to food consumption and specific diets(Thibert,2012).

With minimal food production in cities, food is imported and externalizes the environmental impact of the food urban populations consume. To problematize urban dependence on rural agriculture and ecosystems, this paper investigates the resource support to food preparation and diets in Rosendal,a new residential area currently being developed and touted as a“green”district in Uppsala,Sweden. Apart from evaluating the sustainability of current food consumption patterns including meat,i.e.the actual diet in this case study,the analysis is expanded to examine potential implications of dietary transition, emphasizing changes in the environmental support and sustainability should urban populations shift to a vegetarian(including dairy and eggs)or pescatarian(vegetarian including fish and seafood products) diet. These scenarios are explored because several studies suggest that consuming less animal based foods reduces the environmental impact and support behind food consumption(Garnett,2014;Tilman and Clark,2014;Tom et al.,2016).

A study by Russo et al. (2014)examines the urban metabolism of Uppsala municipality that includes food consumption of both urban and rural populations. Based on their findings, food is the third highest contributor to the amount of input resources that support the urban metabolism of Uppsala (ibid.). Of these, meat has the highest emergy or environmental support followed by fish, oils, dairy and eggs, cereals and derived, fruit and vegetables, and alcoholic beverages (ibid.). The majority of these foods are imported, which requires an expanded systems view when considering the environmental support to food consumption practices and the implications of specific diets.

1.1 Goal of the study

The goal of this study is to assess the resource support to urban food preparation and diets,and compare implications of potential dietary changes. Rosendal is used as a case study representing a modern housing area marketed with an emphasis on green values. Data on the average diet in Rosendal is integrated with emergy synthesis to assess and exemplify the resource support and environmental work behind diets in an urban district officially considered sustainable from an urban planning perspective. Two hypothetical scenarios are developed to explore the potential implications of altered dietary practices of urban populations,and to identify where improvements could be made to food systems sustaining urban populations.

2 Materials and methods

Emergy synthesis is used to evaluate the diet of the population in Rosendal,an urban district located in Uppsala municipality in Sweden. Rosendal was selected as a case study because it is a newly developing district that is promoted as “green” and sustainable. It aspires to become one of the first urban districts in Sweden to be certified sustainable by Sweden Green Building Council(SGBC,2016). Rosendal hence represents a pioneering case of national and municipality promoted sustainable urban development.

While the specific district benefits directly from the scrutiny that this in-depth analysis provides,the results,though primarily case specific, are contextualized and used also to problematize the more general correlation between formal sustainability aspirations in urban planning, environmental performance and support resulting from such policy aspirations. Furthermore, the specific diets evaluated correspond to globally observed trends of urban populations shifting to a western diet. As such, the study is both case specific and representative for dietary transitions currently occurring in urban areas worldwide. In addition, the study evaluates the potential implications of transitioning from diets including meat, to pescatarian or vegetarian alternative diets, which is another aspect of high general relevance for urban planning and policy.

Data on food preparation and diets, including the food and beverages consumed by residents in Rosendal,is sourced from a field study by Maassen (2017). Unit emergy values (UEVs) were obtained from previous publications and databases, as referenced in the emergy evaluation tables. All UEVs used in this study were adjusted to the global emergy baseline 12.0E+24 seJ y-1proposed by Brown et al. (2016). An emergy synthesis was then performed based on the annual diet of the average resident in Rosendal. Subsequently,two hypothetical scenarios were developed to examine if a vegetarian or pescatarian diet would imply a relatively more or less sustainable path for urban populations.

2.1 Characteristics of Rosendal,Uppsala

Rosendal is part of CityLab,a forum for sharing information on sustainable urban development(Uppsala Kommun,2015;SGBC,2016),and is one of twelve urban construction projects in Sweden to be certified sustainable(Uppsala Kommun, 2015). The developers of Rosendal have high sustainability ambitions guided by the keywords of“small-scale” and“diversity” as characteristics for the area(Uppsala Kommun,2015). The vision for the area is proximity and,therefore, to become a mix of residential areas,schools, parks and natural areas, and commercial sites (Uppsala Kommun, 2015). In this urban district, the goal is to achieve as many synergies as possible by promoting multi-functionality, resource efficiency, and to provide safety and comfort for all ages(Uppsala Kommun,2015). However,as in most urban planning and development projects,food is not addressed and Rosendal does not have any explicit food strategy(Maassen,2017). Their view of sustainability emphasizes proximity on a social basis and not the biophysical resource support necessary for urban life.

2.2 Data on the diets in Rosendal

The main input data for this study is derived from Maassen (2017) and was collected through semi-structured interviews guided by questionnaires with 33 participants, including 17 females and 16 males in the ages of 19 to 75. The questionnaires contained a 24-hour recall(c.f. Block,1982)and asked participants to describe their weekly food and beverage consumption in 14 main food categories using a food atlas from the National Food Agency,Sweden(Livmedelsverket,2009). Maassen(2017)collected additional data on the time spent cooking,food waste per week,money spent on food,number of restaurant visits per month,demographics,and household and apartment size(ibid.). The household size and apartment size are used to calculate average kitchen space per capita in the emergy synthesis(ibid). After the semi-structured interviews,Maassen(2017)categorized the food intake reported in the 24-hour recall to the same 14 categories as the weekly estimation. The mean consumption of each food category per day was calculated to represent the average diet of a typical resident in Rosendal, to calibrate for participant over-and under-estimations,then further categorized into 10 broad food categories and extrapolated to monthly values (ibid.). By this strategy, estimated mean values where obtained and assumedrepresentative for the total population in Rosendal. Once the mean monthly food consumption was found for the average resident in Rosendal,this information was used to determine the nutritional value of the diet per day,the cost per month,and the percent each category contributes in terms of total kilograms, total contribution to cost and total kilocalories consumed (Table 1). This data is extrapolated to yearly values for the emergy synthesis.Though diets could fluctuate throughout the year,the diet is assumed constant and representative as an estimated average of the yearly consumption and food preparation.

Table 1 Food consumption of a typical resident in Rosendal represented in kilograms (kg), price per unit in Swedish kronor(SEK),cost or money spent on food in Swedish kronor(SEK),nutritional value in kilocalories(kcal)per day,and the percent of each food category toward total kilograms,cost and kilocalories per month/capita(from Maassen,2017).

2.3 Emergy

In this paper,emergy synthesis is used because,apart from what we eat,it is important to also acknowledge and account for how food is produced(Garnett,2014). Urban food security and diets depend on a range of resources,consumed either directly or indirectly,and include materials,technology,labor,and monetary and energy flows.With emergy,the series of transformation processes behind these resources can be accounted for and expressed in a shared unit: solar emergy joules (sej). Consequently, depending on how specific resources are produced,they attain their own unique position in the global energy hierarchy, a quality expressed as Unit Emergy Value(UEV).Since the same resource can be produced in very different ways,judgments had to be made to identify the most appropriate UEVs for the specific resources evaluated. In this study,and to the extent possible,UEVs from the literature were sourced from other emergy studies in as similar climatic zones and nations as possible then double-checked for accuracy. All UEVs used were adjusted to the global emergy baseline proposed by Brown et al. (2016)and are given in emergy per unit of input(e.g. sej/J,sej/g,sej/SEK).

Prior to the emergy synthesis, the systems boundary, i.e. window of attention, and inputs were defined and described through a system diagram. Subsequently, the data for the diets and food preparation in Rosendal obtained from Maassen(2017),including all material,money and energy flows,were converted into solar emjoules and represent the environmental support behind the annual diet and food preparation per capita (Table 2). Additional labor emergy exceeding the portion fed back through metabolic emergy was disregarded in this study when calculating emergy for time spent cooking. After determining the total solar emergy and outputs of the system,two hypothetical scenarios were developed based off of suggestions from previous research.

Indices and ratios were calculated to evaluate the renewability and sustainability of the average diet in Rosendal as well as the hypothetical scenarios. The following indices and ratios were used to evaluate and compare the scenarios: emergy yield ratio(EYR,Y/F),percent renewable(%Renew,R/Y),environmental loading ratio (ELR, (F+N)/R), emergy sustainability index (ESI, EYR/ELR), and solar cost index (SCI, Y/solar share). See Fig. 2 and table 6 for details on how calculations were performed in this study, and Odum(1996)for a comprehensive introduction to emergy accounting.

2.4 Assumptions for the scenarios

The hypothetical scenarios include a vegetarian(including eggs and dairy)and pescatarian diet(vegetarian diet that includes fish and seafood). These diets are to evaluate the potential implications of alternative diets, with regards to environmental support and total emergy. These diets are selected based on previous research that suggest consuming less animal products reduces the environmental impact and support behind a given diet(Garnett,2014;Tilman and Clark,2014;Tom et al.,2016).

For the vegetarian scenario,it is assumed that all the meat and fish consumption is substituted by,and hence apportioned to, the vegetable and fruit category given the aggregation of legumes with the fruit and vegetable category in the Maassen (2017) study. To account for the differences in nutrition derived from these different food products,the amount of meat and fish consumed was multiplied by the amount of fruits and vegetable that are needed to supplement the nutrition that is no longer obtained from meat and fish. The ratio of fruit and vegetables necessary to substitute meat and fish used in the calculation is based on Gilsing et al. (2013). As for the pescatarian diet, fish consumption is assumed to remain constant and the amount of meat consumed is assumed to be substituted by more fruits and vegetables as in the vegetarian scenario. In both scenarios, the money expended to procure food is assumed to remain constant.

3 Result s

3.1 Food preparation and diets

Table 1 presents the food consumption data from Maassen (2017) in mean monthly kilograms (kg) per capita seen under amount as well as the cost per unit and per month, the nutritional value and percent contribution of each food category, amount of kilograms, cost and kilocalories (kcal). The cost per unit is based on the prices used by Russo et al. (2014), which gives a total monthly food expense of 1479 SEK. The average person in Rosendal consumes approximately 56 kg of food in a month with fruit and vegetables as the largest contributor to total weight (23%), which is tied with meat for the highest cost per month. Combined, meat and fruit and vegetables constitute 38% of monthly food expenses. In terms of kcal, cereals and derived have the highest energy yield followed closely by dairy and eggs then fruit and vegetables. Fats also contribute a significant amount in kilocalories in relation to their contribution to total kilograms or cost. Though,a significant contribution to both total kilograms and kilocalories is from non-alcoholic beverage consumption. On the other hand, meat and fish have relatively low nutrient and total kilogram contribution. Even with the low kilogram quantity,meat and fish are some of the more expensive food items consumed. Similarly,stimulants contribute a noteworthy amount in terms of kilograms yet have no nutritional value. Therefore, the diet of a typical average Rosendal resident receives relatively little nutrients from stimulants,fish,alcoholic beverages and meat yet these comprise 39%of monthly costs and 29%of monthly kilograms. Still,food consumption in Rosendal is on the high end of the English National Health Service (NHS, 2016) daily recommendation of 2000-2500 calories whereas the municipal average is on the lower end.

3.2 Emergy synthesis

The food system diagram defines the window of attention or system boundaries as the kitchen where the food is prepared, its main inputs and generated outputs (Fig. 1). According to systems theory and emergy diagramming conventions, the inner frame in Fig. 1 is the window of attention of this study, thus indicating processes internal to the system. The outer frame represents the next larger scale,the embeddedness in Rosendal,Uppsala municipality,and in extension the global food system.

Fig.1 Food systems diagram for food preparation and consumption in Rosendal.

As seen in Fig.1, imported resources support the internal transformation process of food preparation, generating a food storage that is consumed by the residents. The process is enabled by drawing from a storage of money to pay for the services embedded in producing, processing, packaging and delivering the goods to Rosendal. The inputs are defined in terms of three types of resources: local renewable (R),imported resource inputs (F), and the system yield (Y). From the inputs and processes within the system, outputs are generated and seen as outflows(Y)on the right of the diagram. The two outputs generated from this system include food waste and the human metabolism required for residents to live, work and otherwise participate in society. The outputs are seen as two inherently different co-products that share the same emergy inputs and the time and labor dedicated to cooking is considered a feedback within the system.

Resources external to the system support the diet and food preparation in a variety of ways. The local renewable inputs(R)refer to the resource inputs that are regenerative from outside the system boundaries. Local nonrenewable inputs(N)are not regenerative within the system boundaries,and imported inputs(F)are brought into the system from the larger economic production system. Since all resource inputs had to be imported to Rosendal,there are no local nonrenewable inputs in this system as shown in Fig. 1. Fig.1 combines all resources and components that support food preparation and consumption in Rosendal to show the inputs, outputs, and interactions between these. In addition, Fig. 2 effectively defines how the input flows were aggregated and grouped to calculate the emergy indices and ratios.

To capture the percent renewable embedded in the purchased imported inputs (F),the window of attention is expanded to include the percent renewable in F inputs to produce F2 and R2, see Fig. 3. R2 is defined as solar insolation and the renewability fraction of all the system inputs. The remaining inputs are defined as F2 and include the nonrenewable fraction of electricity,appliances,the physical structure,human services, and the food and beverage inputs. Human services are an important input category in emergy syntheses and express the indirect environmental and economic support received through imports (feedback). In this case, services accounted for are those embedded in kitchen appliances,building depreciation,purchased food items,electricity,and water (Tables 2, 3 and 4). This not only takes historical emergy into account but also the resources used in production, processing, packaging, distribution in addition to the human labor and knowledge behind these inputs. Our system has no N and N2is not calculated specifically but is included in F2.

For all three scenarios, there are 138 kilograms of food waste per year and 353 hours of labor (Maassen,2017). From these outputs,new UEVs are calculated based on the output quantities and the total emergy(Table 2, 3,and 4). For the vegetarian and pescatarian scenarios, the imported emergy flows are adjusted to represent food consumption based on these types of diets,as described in section 2.4.

Fig.2 Aggregated emergy flow diagram used to define and calculate flow categories into renewable(R),nonrenewable(N)and purchased(F)emergy entering the system and its emergy yield(Y)in Tables 2,3,and 4. Calculations are also shown for the different emergy indices.

Fig. 3 Emergy systems diagram showing that the purchased resources F also are fed by a portion of renewable and nonrenewable emergy. Information we used when we extend the window of attention in order to calculate the renewable portion of emergy feeding the diets.

The total emergy supporting the diets including food preparation is the highest for the actual diet of the typical resident in Rosendal. The lowest total emergy is associated with the vegetarian scenario,but the pescatarian diet is comparable in total emergy and lower than the actual diet. The environmental support for the vegetarian scenario is the least intensive at 1.65E+16 sej followed closely by the pescatarian scenario at 1.69E+16 sej. For all scenarios, the single largest emergy input is F,human services (item 15), at 1.51E +16 sej. The remaininginputs are one to seven orders of magnitude less than total human services.

Table 2 Emergy support to food preparation and the actual diet per capita/year in Rosendal,Uppsala.

Table 2 Continued.

The generated outputs from all scenarios are waste and human metabolism. UEVs for these outputs are calculated by dividing the total emergy,including and excluding services, by the remaining available energy of the outputs in solar emjoules. When including services,calculated UEVs are an order of magnitude higher than when excluding services for all three scenarios.

3.3 Emergy indices and ratios

The emergy indices and ratios were calculated using the values seen in Table 5 to facilitate comparison and understanding of the renewability and relative efficiency of the scenarios(Table 6). Slight variations are evident in the indices and ratios when comparing the three diet scenarios,but all scenarios have similar values.

Systems running on high shares of renewable inputs are more likely to be sustainable than those with low percent renewable. The actual diet, including indirect (imported) renewable flows and services, has a total percent renewable of 3.20%. This is higher than the vegetarian diet(2.34%),but lower than the pescatarian diet(3.24%). When excluding services,the percent renewable is highest for the pescatarian diet at 23.84%renewable followed by the vegetarian diet at 18.48% and subsequently the actual diet at 15.25%. The difference derives from the fact that all scenarios are predominantly dependent on indirect emergy in the form of services, which implies that excluding these results in significant changes to the total renewability. It is also partly explained by the fact that the average renewability of services in Sweden,which was used in this study,is only 0.8%percent(NEAD,2008). However, this discrepancy does not apply to the emergy sustainability index (ESI),which forall three scenarios is very low by global standards. The ESIs for all three scenarios are low with values of 0.03 for both the actual and pescatarian diets and 0.02 for the vegetarian diet.

Table 3 Emergy support to food preparation and a vegetarian diet scenario per capita/year in Rosendal,Uppsala.

Table 4 Emergy support to food preparation and a pescatarian diet scenario per capita/year in Rosendal,Uppsala.

Table 5 Aggregated renewable and imported flows and total emergy yield for the diet scenarios. Flow categories labeled R1,F1,and Y are calculated as suggested by Brown and Ulgiati(1997).Categories labeled R2 and F2 distinguish between the renewable and nonrenewable components of emergy inputs,i.e. where the window of attention has been expanded to acknowledge contribution of indirect renewable emergy(services included).

Table 6 Calculated indices derived from the aggregated flows in Table 5. Original expressions from Brown and Ulgiati(1997).

To estimate the relative stress these scenarios place on the environment, the environmental loading ratio(ELR)is used. ELR is the ratio of nonrenewable and imported inputs to renewable inputs. The vegetarian diet has the highest environmental load of 41.97 followed by the actual diet at 30.28. The pescatarian diet has the lowest ELR of 29.88. However,all scenarios perform poorly if considering only the ELR,which can be expected from systems that depend more on imported inputs than local renewable inputs. The results are similar when interpreting the solar cost index(SCI).The actual diet appropriates the largest solar(emergy)share as indicated by the high solar cost index(SCI).A SCI of 11.09 for current diet indicates that the average resident in Rosendal consumes 11.09 times its fair solar share based on its diet alone. Were the average resident in Rosendal to switch to a vegetarian or pescatarian diet,the SCI would decrease to 10.13 or 10.35 times its fair share,respectively.

4 Analysis and discussion

This case study shows that the human metabolism of urban populations is primarily a consumer system. The total emergy support to the diets in Rosendal is 1.81E+16 sej/capita/year,a number that surpasses the municipal average by a magnitude of five compared to the previous estimate by Russo et al. (2014). This means that the district of Rosendal places more stress on its system and is less sustainable than the municipal average. Some of the increases in emergy could be explained by relatively higher consumption of meat,fats,and dairy and eggs in comparison to the municipal average,which includes rural populations. These findings correspond to the trends seen between rural and urban populations shown by Kearney(2010),for example.

Shifting to a vegetarian diet would decrease the total emergy support to 1.65E+16 sej/capita/year whereas a pescatarian diet would imply a change to 1.69E+16 sej/capita/year. The lower total emergy is explained by the conversion of the meat input to fruits and vegetables since meat has the highest contribution to total emergy when excluding services.

For the vegetarian diet,the overuse of its solar share is the lowest but at the cost of an even lower renewability.The pescatarian diet in Rosendal has the highest percent renewable when including indirect renewability and excluding services at 23.84%, which is partly based on the next larger system due to the expanded window of attention. The direct percent renewable is nearly the same for each diet scenario when services and indirect R are included, ranging within 1%from 2.34%for the vegetarian scenario to 3.24%for the pescatarian scenario.This discrepancy can be explained by the differences in the total emergy in relation to services since the cost per month is constant. However, regardless of the calculation procedure chosen, the ELR ratio is high in all scenarios. With the strong reliance on imported inputs,challenges are present in the ability to generate feedbacks within the system. This is also indicated by the low EYR,which means the system does not efficiently utilize local renewable resources. From a different perspective, the low EYR may be seen as an underused potential meriting further exploration.

In terms of the general distribution of total emergy,services account for the single largest input category, at 83.41% of the total solar emergy support to the actual diet. In this scenario, the second highest solar emergy input is meat, which contributes 6.87% of the total emergy. These ratios change, however, for the two alternative scenarios. In the vegetarian and pescatarian scenario, the contribution of services amounts to 91.31% and 89.42%,respectively. The change is in part because of the reduced total emergy and the conversion of the meat input into fruits and vegetables.

Excluding services,meat has the highest solar emergy and has the highest contribution of all other remaining inputs. Previous studies also show that meat has the highest impact on the environment as well as GHG emissions in comparison to other food products(Scarborough et al.,2014;Tilman and Clark,2014). When comparing meat with the other food products,its impact is also particularly evident from an emergy perspective. When excluding services and the physical structure,meat contributes 29%more emergy to the system than the next highest food input and double the emergy of the next highest input (physical structure). However, in the pescatarian and vegetarian diet, when excluding services and the physical structure, the second highest contributor is fish and fats,respectively. This also aligns with previous research on urban populations and a shift to a western diet. The contribution of these inputs in the diet scenarios is in part due to having some of the highest UEVs after services.

Based on the three diet scenarios, imported inputs (F) therefore amount to nearly all of the solar emergy supporting the diets in Rosendal which overshadows all other inputs. In this system, there are no N inputs since the local, nonrenewable resources have been imported. The local renewable inputs (R) almost have no contribution to the system when the percent indirect renewable is excluded from R.For example, R-inputs for the actual diet only contribute 0.000017%of the solar emergy to the system. Therefore,in this paper,the percent indirect renewable has been separated from the F inputs and accounted for as indirect R,and represented in the indices and ratios as R2.

The outputs from the system in all scenarios are waste and human metabolism. Differences in the UEVs are a result of the variations in total emergy supporting each of these different scenarios. The high UEVs for each scenario however still imply a high quality resource resulting from a high resource throughput system. The application of these UEVs to social life in other systems is suggested to be used with careful consideration,because these values only represent food preparation and the average diet in Rosendal and human life requires many more resources than food.

5 Conclusions

This study contributes several new UEVs, calculated in different ways to enable comparison and use in other studies. The UEVs also provide an understanding of where Rosendal,as a green urban district,stands in terms of food system sustainability. The results show that Rosendal is a consumer and throughput system that is neither efficient nor sustainable. The environmental support to the average resident’s food preparation and consumption in Rosendal is significant at 1.81E+16 sej for the current diet,1.65E+16 sej for the vegetarian diet,and 1.69E+16 sej for the pescatarian diet. This demonstrates the importance of considering food preparation and diets both when assessing and planning for sustainable urban development. The results indicate that the diet itself is not necessarily the cause of poor performance, and that the larger system of production and distribution is more influential. Therefore, new models of production and consumption are needed for a sustainable autocatalytic food system.

The results also show that the typical resident no matter the diet, overshoots the solar share strictly based on their food preparation and consumption alone. This challenge is further aggregated considering that both the municipal and national averages are already higher than the global average (Johansson, 2005; Russo et al.,2014). From a policy perspective,this could be problematic since the results are conflicting with Uppsala’s 2050 vision of developing towards a municipality that operates within planetary ecological limits(Uppsala Kommun,2015). Furthermore,the analysis is delimited to food preparation and diets though additional emergy is required to sustain other crucial activities and resources for human life(e.g. clothing,shelter,transportation, etc.). Thus,it may be expected that an expanded analysis of the resource use associated with all other aspects of urban life would significantly increase the degree to which urban lifestyles overshoot their fair share of available global emergy.

To improve sustainability in Rosendal,as well as in other food systems around the world,a shift is required from consuming foods derived from the global food system to foods produced in systems that generate autocatalytic feedbacks and source inputs from local renewable resources. Likewise, reducing meat consumption could also be beneficial to reduce greenhouse gas emission as well as the overall efficiency of the food system.This study suggests that it is unlikely to achieve sustainable urban development without the consideration of the global food system in its entirety and in parallel to other urbanization trends and challenges facing cities today and in the future. Though,this study does not consider instances where food is sourced from(peri-)urban food production or where cities are supplied with food primarily from their rural hinterlands through community supported agriculture, for example. In these contexts, urban food systems may not be as reliant on the global food system as is the case in our study.

Acknowledgements

The work presented in this paper was funded by a grant from the Swedish Research Council for Environment,Agricultural Sciences and Spatial Planning (Formas). Additional support was provided by the Department of Urban and Rural Development,Swedish University of Agricultural Sciences(SLU).