University of Groningen Natural resource use for food Leenes ...Wiskunde en Natuurwetenschappen aan de Rijksuniversiteit Groningen op gezag van de Rector Magnificus, dr. F. Zwarts,

University of Groningen

Natural resource use for foodLeenes, Popkje Winfrieda

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2006

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Leenes, P. W. (2006). Natural resource use for food: land, water and energy in production andconsumption systems. s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 21-08-2021

https://research.rug.nl/en/publications/natural-resource-use-for-food(ff10ce89-8f88-42da-8355-99f3bd806e59).html

NNaattuurraall rreessoouurrccee uussee ffoorr ffoooodd:: llaanndd,, wwaatteerr aanndd eenneerrggyy iinn pprroodduuccttiioonn

aanndd ccoonnssuummppttiioonn ssyysstteemmss

ii

Copyright © by P.W. Gerbens-Leenes ISBN 90-367-2868-1 Printed by Facilitairbedrijf, Groningen, the Netherlands Cover: the Lienesch in Ankum (Germany), photo by Bauke de Vries

iii

RIJKSUNIVERSITEIT GRONINGEN

NNaattuurraall rreessoouurrccee uussee ffoorr ffoooodd:: llaanndd,, wwaatteerr aanndd eenneerrggyy iinn pprroodduuccttiioonn aanndd

ccoonnssuummppttiioonn ssyysstteemmss

NNaattuuuurrlliijjkkee hhuullppbbrroonnnneenn vvoooorr vvooeeddsseell:: llaanndd,, wwaatteerr eenn eenneerrggiiee iinn pprroodduuccttiiee-- eenn ccoonnssuummppttiieessyysstteemmeenn

Proefschrift

ter verkrijging van het doctoraat in de Wiskunde en Natuurwetenschappen aan de Rijksuniversiteit Groningen

op gezag van de Rector Magnificus, dr. F. Zwarts, in het openbaar te verdedigen op

vrijdag 15 december 2006 om 14.45 uur

door

Popkje Winfrieda Leenes

geboren op 17 september 1953

te Groningen

iv

Promotor: : Prof.dr. A.J.M. Schoot Uiterkamp : Prof.dr. H.C. Moll Copromotor : Dr.ir. S. Nonhebel Beoordelingscommissie : Prof.dr. J. Van Andel : Prof.dr. K. Blok : Prof.dr. R. Leemans

v

Everything in excess is opposed to nature.

Hippocrates

vi

vii

VVoooorrwwoooorrdd Dit proefschrift kent een lange voorgeschiedenis die begon met mijn studie diëtetiek in Nijmegen in 1971. Daar kwam ik voor het eerst in aanraking met het vakgebied van de voedingswetenschappen. Maar, behalve theoretische kennis, werd mij daar ook de kookkunst in de praktijk bijgebracht. Jaren later, het was inmiddels 1991, begon ik mijn studie milieuwetenschappen aan de Open Universiteit met het idee iets heel anders te gaan doen dan voedingswetenschappen. Omdat de Open Universiteit geen onderzoeksfaciliteiten had in mijn richting, zocht ik voor mijn scriptie een plaats aan een “gewone” universiteit. Dat werd de Rijksuniversiteit Groningen. Bij het Centrum voor Energie en Milieukunde IVEM was men al geruime tijd bezig met onderzoek naar de milieueffecten van huishoudelijke consumptie, waaronder…..voeding. Die scriptie ging dan ook over de relatie tussen land, energie en Nederlandse huishoudelijke voedselconsumptie. Na het afronden van de scriptie en het behalen van mijn doctoraal in de milieuwetenschappen, bleef ik bij de IVEM en schreef daar het “Groene Kookboek”, waarin de voedings- en milieukundige kennis werden vertaald voor een groter publiek. Toen kort daarna een aio-plaats beschikbaar kwam voor onderzoek naar milieukundige effecten van voeding, kon mijn promotietraject beginnen. De periode van onderzoek heb ik als een geweldige tijd ervaren. Op de eerste plaats vanwege de stimulerende omgeving waarin ik verkeerde. Niet alleen ben je met je eigen onderzoek bezig, maar ook dat van anderen komt ruimschoots aan bod, zowel binnen als buiten de eigen universiteit. Zo waren daar de cursussen en de bijeenkomsten van de onderzoeksschool SENSE, een deelname aan een workshop op Sicilië, een zomerschool in de Franse Alpen, en een verblijf van drie maanden bij IIASA in Wenen in de zomer van 2004. Tijdens mijn promotietraject heb ik veel steun gehad. Op de eerste plaats wil ik mijn eerste promotor, Ton Schoot Uiterkamp noemen. Toen ik een afstudeerplek zocht voor mijn scriptie zag hij wel wat in die studente van de Open Universiteit. Ook voor de invulling van de aio-plaats in 2000 had hij het volste vertrouwen, dat hij met de keuze voor mij de juiste persoon op de juiste plaats had gezet. Met de voltooiing van mijn proefschrift hoop ik te hebben laten zien dat het vertrouwen van Ton in mij terecht was. Voor mijn tweede promoter, Henk Moll, geldt eveneens dat ik veel te danken heb aan zijn vertrouwen en wijze begeleiding tijdens het promotietraject. Een bijzondere plaats is ingenomen door Sanderine Nonhebel. Niet alleen was ze mijn dagelijks begeleidster, maar ook mijn kamergenote. Mijn ontwikkeling tot zelfstandig onderzoeker is voor een groot deel haar werk. Niet alleen wist ze mij de kneepjes van het vak dat wetenschap heet bij te brengen. Bovendien heb ik van haar ook geleerd artikelen te schrijven. Behalve van mijn begeleiders, heb ik ook veel steun gehad van al mijn andere IVEM collegae, die mij hebben bijgestaan met alle vragen die ik zoal had. Bedankt Sandra Bellekom, René Benders, Michiel Berger, Dick van den Berg, Emiel Elferink, Michiel Hekkenberg, Laurie Hendrickx, Annemarie Kerkhof, Sander Lensink, Nicole van Marle, José Potting, Niels Schenk, Anne Jelle Schilstra en Frauke Urban. Een bijzondere stimulans kwam van mijn familie. Zonder hun steun was dit proefschrift nooit tot stand gekomen. Als klankbord bij alle promotieperikelen heeft vooral Sepp mij met raad en daad bijgestaan. Met veel geduld heeft hij mijn teksten van correcties voorzien. De eerste schrijfsels kreeg ik vaak meer “rood” dan “zwart” terug. Terugziend op het hele traject ben ik er van overtuigd geraakt dat je succes maar gedeeltelijk kunt afdwingen. Mede dankzij alle genoemden is mij dat overkomen.

viii

ix

CCoonntteennttss

VVoooorrwwoooorrdd VViiii

11 GGeenneerraall iinnttrroodduuccttiioonn 11

1

1

2

1.1 Introduction

1.2 Global food issues

1.3 Food systems

1.4 Food requirements 2

3

1.4.1 Dietary requirements for the basic and subsistence level

1.4.2 Food consumption patterns 3

3

3

1.5 Food and sustainability

1.6 Scope of the thesis

1.7 Structure of the thesis 5

22 DDeessiiggnn aanndd ddeevveellooppmmeenntt ooff aa mmeeaassuurriinngg mmeetthhoodd ffoorr

eennvviirroonnmmeennttaall ssuussttaaiinnaabbiilliittyy iinn ffoooodd pprroodduuccttiioonn ssyysstteemmss

77

Abstract 7

2.1 Introduction 7

2.2 System description 9

2.3 Methods 11

2.3.1 Analysis of existing measuring methods and scientific research

issues

11

2.3.2 Design and development of an environmental measuring method 11

2.4. Results 11

2.4.1 Environmentally sustainable business practices 11

2.4.2 Sustainability indicators from the perspective of companies 12

2.4.3 Sustainability issues from the perspective of environmental

research

13

2.5. Design and development of a measuring method 13

2.5.1 Integration of bottom-up and top-down approaches 14

2.5.2 Selection of indicators 14

2.5.3 Flow chart of calculations and input 16

2.6. Discussion 18

2.6.1 Present situation: type of information generated and utility 18

2.6.2 The measuring method: type of new information and utility 19

2.6.3 Future research: type of additional information required 20

2.7 Conclusions 20

x

33 AA mmeetthhoodd ttoo ddeetteerrmmiinnee llaanndd rreeqquuiirreemmeennttss rreellaattiinngg ttoo ffoooodd

ccoonnssuummppttiioonn ppaatttteerrnnss

2211

Abstract 21

3.1 Introduction 21


3.3 Materials and methods 23

3.3.1. Scale levels for land requirements for food 23

3.3.2. Flow chart of calculations and input 24

3.4. Results and discussion 26

3.4.1 Land requirements 26

3.4.2 Effect of uncertainty and inaccuracy on final results 28

3.4.3 Comparison with data on available food on a national level 29

3.4.4 Application and sensitivity of the method 29

3.5 Conclusions 30

44 CCoonnssuummppttiioonn ppaatttteerrnnss aanndd tthheeiirr eeffffeeccttss oonn llaanndd rreeqquuiirreedd

ffoorr ffoooodd

3333

Abstract 33

4.1 Introduction 33

4.2 Food consumption patterns 35


4.3.1 Starting points 35

4.3.2 Land requirements for basic and subsistence consumption 36

4.3.3 Land requirements for culturally defined consumption patterns 36

4.3.3.1 Inter-generational differences 37

4.3.3.2 Regional differences 37

4.3.4 Presentation of results 37

4.4 Results 38

4.4.1 Land requirements for the basic and subsistence levels 38

4.4.2 Consumption and related land requirements for the cultural level 38

4.4.2.1 Dutch inter-generational differences 38

4.4.2.2 Regional differences 40

4.4.2.3 Correction for energy intake 43

4.5 Discussion 43

4.6 Conclusions 44

xi

55 CCrriittiiccaall wwaatteerr rreeqquuiirreemmeennttss ffoorr ffoooodd,, mmeetthhooddoollooggyy aanndd

ppoolliiccyy ccoonnsseeqquueenncceess ffoorr ffoooodd sseeccuurriittyy

4477

Abstract 47

5.1 Introduction 47


5.2.1 Food crops and their place in human nutrition 49

5.2.2 Crop production 49

5.2.3 Water flows at a crop field 50


5.3.1 Starting points 51

5.3.2 Hypothetical crops as representatives for crop types 52

5.3.3 Radiation use efficiency and glucose for growth 52

5.3.4 The transpiration of water 52

5.3.5 ‘Transpirational’ water requirements 53

5.4 Results 53

5.5 Discussion 54

5.5.1 Applicability 54

5.5.2 Options to reduce water requirements by changing food

consumption patterns

54

5.5.3 Increasing global food production 57

5.6 Conclusions 57

66 FFoooodd ccoonnssuummppttiioonn aanndd eeccoonnoommiicc ddeevveellooppmmeenntt,, aa ssppaattiiaall

aanndd tteemmppoorraall ccoommppaarriissoonn

5599

Abstract 59

6.1 Introduction 59

6.2 Food systems 60

6.2.1 Agricultural production 61

6.2.2 Food industry 62

6.2.3 Household consumption 62

6.2.4 Per capita consumption 62

6.2.5 Physical streams in the food system 62

6.3 Section l, spatial differences among food supply and consumption 62

6.3.1 Introduction Section l 62

6.3.2 Materials and methods Section l 63

6.3.2.1 Units of calculation 63

6.3.2.2 Income, food supply and per capita consumption 63

6.3.3 Results and discussion Section l 64

6.3.3.1 Per capita income and food supply 64

xii

6.3.3.2 Per capita income, composition of consumption and

contribution of animal foods

64

6.3.3.3 Uncertainty and inaccuracy of results 66

6.3.4 Conclusions Section l 67

6.4 Section ll, the use of supply data 67

6.4.1 Introduction Section II 67

6.4.2 Materials and methods Section ll 67

6.4.3 Results and discussion Section ll 68

6.4.3.1 The composition of supply and consumption 68

6.4.3.2 Per capita income and nutritional energy intake 68


6.4.4 Conclusions Section ll 69

6.5 Section lll, temporal differences among food supply and consumption 69

6.5.1 Introduction Section lll 69

6.5.2 Materials and methods Section lll 69

6.5.3 Results and discussion Section lll 69

6.5.3.1 Food supply in France and Great Britain, 1700-2000 69

6.5.3.2 Food supply and consumption in southern Europe, 1961-

2000

70

6.5.3.3 Comparison of results with information from food surveys 70


6.5.4 Conclusions Section lll 73

6.6 General discussion 73

6.6.1 Trends 73

6.6.2 Future changes 73

77 PPaatthhwwaayyss ttoowwaarrddss ssuussttaaiinnaabbllee ffoooodd ccoonnssuummppttiioonn ppaatttteerrnnss 7755

Abstract 75

7.1 Introduction 76

7.2 Section l, natural resource use for food consumption patterns in time

and space

77

7.2.1 Introduction Section l 77

7.2.2 Materials and methods Section l 77

7.2.2.1 ‘Transpirational’ water requirements for foods 77

7.2.2.2 Land, ‘transpirational’ water, and energy for the Dutch

food consumption pattern

77

7.2.2.3 Long-term trends in actual land, water and energy

requirements

78

7.2.2.4 Resource use in a developing and in a developed

country

78

xiii

7.2.3 Results and discussion Section l 78

7.2.3.1 ‘Transpirational’ water requirements 78

7.2.3.2 Resource use for the Dutch food consumption pattern in

1990

80

7.2.3.3 Long-term trends for land, water and energy of the Dutch

food consumption pattern

82

7.2.3.4 Poor and affluent food consumption patterns 85

7.3 Section ll, pathways towards sustainable food consumption patterns 85

7.3.1 Introduction Section ll 85

7.3.2 Materials and methods Section ll 85

7.3.2.1 Increased efficiency of production 85

7.3.2.2 Prevention 85

7.3.2.3 Substitution 87

7.3.3 Results and discussion Section ll 87

7.3.3.1 Options for increased efficiency 87

7.3.3.2 Options for prevention 88

7.3.3.3 Options for substitution within food categories 88

7.3.3.4 Options for substitution among food categories 89

7.4 General discussion 90

7.5 General conclusions 91

RReeffeerreenncceess 9933

GGlloossssaarryy 110077

AAppppeennddiixx AA Land requirements, indirect energy requirements and household

requirements for food items in the Netherlands in 1990

110099

AAppppeennddiixx BB Overview of the fifty two countries for which Chapter 6 performed

a spatial analysis

111177

AAppppeennddiixx CC OOverview of countries with national food surveys used in Chapter

6 and their authors

111199

SSuummmmaarryy 112233

1. Introduction 123

2. Food systems 123

3. Measuring method for environmental sustainability in food production

systems

123

4. Method to determine land requirements for food consumption patterns 124

xiv

5. Consumption patterns and land required for food 125

6. Consumption patterns and water required for food 125

7. Food consumption and economic development, a spatial and temporal

comparison

126

8. Pathways towards sustainable food consumption patterns 127

SSaammeennvvaattttiinngg 112299

1. Introductie 129

2. Voedselsystemen 129

3. Ontwerp en ontwikkeling van een meetmethode voor milieukundig duurzaam

ondernemen in de voedingssector

130

4. Methode voor de berekening van het landbeslag voor voedsel 130

5. Voedselconsumptiepatronen en landbeslag 131

6. Voedselconsumptiepatronen en de behoefte aan water 131

7. Voedselconsumptiepatronen en economische ontwikkeling 132

8. Ontwikkeling van duurzame voedselconsumptiepatronen 133

CCuurrrriiccuulluumm vviittaaee 113355

LLiisstt ooff ppuubblliiccaattiioonnss 113377

AAcckknnoowwlleeddggeemmeennttss 113399

1

CChhaapptteerr 11

GGeenneerraall iinnttrroodduuccttiioonn∗∗

11..11 ►► IInnttrroodduuccttiioonn Humans need food to remain healthy and to stay alive. The food derives from production systems that claim a large share of available natural resources, such as land, fresh water, and energy carriers. The main objective of this thesis is to give insight into the relationship between per capita food consumption patterns and the use of natural resources. Differences in the use of these resources among consumption patterns might contribute to the development of more sustainable food consumption. The assessment of the impact of food consumption on natural resources requires information on production systems and their output in the form of commodities and individual food items. This chapter starts by identifying important global food issues. Next, it shows the components of food systems, it describes human food requirements, the impact of food consumption on resource use, and it presents the complex relationship between food and sustainability. The chapter concludes with the presentation of the scope and the objective of the thesis, the central research question, and the outline. 11..22 ►► GGlloobbaall ffoooodd iissssuueess The World Food Summit of 1996 (FAO, 1996) reminded the world once again of the importance of food security and of the fact that the absolute number of food-insecure people is still growing. An adequately nourished population is essential for the absorption of cognitive and vocational skills and forms an important dimension of human capital (Schuh, 2000). Sufficient food of adequate quality available for all humans on this planet is a constraint for sustainable development (Annan, 2002). Four factors dominate food security in the coming decades: (i) the growth of the world population; (ii) changes in agriculture and the food industry; (iii) a shift from local self sufficiency towards a global commodity market; and (iv) changes in food consumption towards more affluent patterns. Especially in developing countries, the growth of the world population is large and requires enormous efforts from agriculture (Tilman et al., 2002). By 2050, the United Nations’ medium projection estimates global population to be 50% larger than in 2003 (United Nations Population Division, 2002). Up to now, changes in agriculture, the second factor, and especially the increase of production has been sufficient to meet the growth of demand (FAO, 2003A). The second World War marked a turning point in the yield per hectare of arable crops in the Western world. For example, before World War ll, yields of wheat in the United Kingdom and the USA increased only by a few kg ha-1 year-1 (De Wit, 1992). Since then, yields have increased consistently at much higher rates. This ‘first green revolution’ was due to a rapid increase in demand. In some countries, the physiological limits of yields are almost reached, however. A continuation of the increase in demand, therefore, implies huge challenges on the availability of land resources. In the coming decades, arable land expansion will be small (FAO, 2003A). Moreover, continuing land degradation and the impact of environmental awareness of consumers puts a strain on the ability to produce enough food (Bouma et al., 1998). The third factor, the shift toward a world commodity market, not only implies a larger availability on a global level (Ivens et al., 1992), but also shifts in the specific availability of commodities. Over time, families in traditional societies have learned what constitutes an adequate diet by trial and error. In an economy that is changing rapidly towards a global market, knowledge about nutrition forms an important constraint for food security (Schuh, 2000). Inuit peoples, for example, consume raw meat and in this way solely provide for their ascorbic acid, an essential vitamin for humans (Receveur et al., 1997).

∗ Sections 1.2 – 1.4 are derived from Gerbens-Leenes, P.W, Nonhebel, S., 2005. Food and land use. The influence of consumption patterns on the use of agricultural resources. Appetite 45, 24-31.

Chapter 1 __________________________________________________________________________________________

2

Food consumption patterns are repeated arrangements that can be observed in the consumption of food by a population group (Ivens et al., 1992; Whitney and Rolfes, 1999). Changes in food consumption patterns, the fourth factor, can be large and sometimes take place in a short period of time. In many countries since 1961, increasing affluence has often gone along with an increase of the consumption of affluent foods like meat and alcoholic beverages (FAO, 2004A). The four factors mentioned above all influence the amount of natural resources, such as agricultural land, fresh water, and energy carriers, required for food. This thesis puts most emphasis on the fourth factor and assesses the influence of food consumption patterns on the use of natural resources but gives also attention to agriculture and the food industry. 11..33 ►► FFoooodd ssyysstteemmss This thesis divides food systems into two subsystems: (i) a production subsystem and (ii) a consumption subsystem. Figure 1.1 shows a simplified conceptual framework showing the factors that determine total, natural resource requirements for food. Total food consumption is determined by the size of the population and the amounts and types of foods that are consumed, i.e. food consumption patterns. The subsystems are quite complex, however, and, moreover, often show a lack of transparency of physical streams. The production subsystem comprises primary and secondary production, as well as the food industry. Primary production grows agricultural crops. Yields per hectare depend heavily on the type of system applied, leading to large variation with regard to yields. For example, average wheat yields in the Netherlands around 1900, when little fertilizer was used and cereals were grown in rotation with legumes, were about 2.0 Mg ha-1 (Spiertz et al., 1992). In 1995, Dutch wheat yields have risen to 8.7 Mg ha-1 (FAO, 1999). Crops from the primary production systems, such as soybeans, barley, or maize, form the basis for the secondary or livestock production.

Fig. 1.1. Conceptual framework showing factors that determine the total resource requirement for food. The food industry processes commodities and manufactures an enormous variety of foods using a large number of ingredients. To manufacture a cake, for example, industry needs a variety of basic ingredients such as sugar, flour, eggs, and butter, and these all have different resource requirements. Therefore, resource requirements for single food items can differ considerably. 11..44 ►► FFoooodd rreeqquuiirreemmeennttss Food has three functions. First, the basic food function is to provide enough nutritional energy for body functions and physical activity. Second, to provide health, food should confirm to nutritional constraints and contain sufficient amounts of vitamins and minerals. Third, food should also meet cultural and emotional requirements. As a result, food packages also contain foods low in nutrient density and more foods than needed to stay alive. The actual food consumption patterns are of course to be found on the cultural level. Different food functions are related to different resource requirements.

Agricultural production and food

industry Human

consumption

TOTAL

RESOURCE REQUIREMENT FOR

FOOD

Population size

Consumption pattern

General introduction

__________________________________________________________________________________________

3

For the three food functions, the thesis distinguished three scale levels (the basic, the subsistence and the cultural level), which are briefly discussed below. 1.4.1 Dietary requirements for the basic and subsistence level The basic function of food is to provide enough energy for body functions and physical activity. To provide sufficient nutritional energy, about 10 MJ per capita per day is needed (Voedingscentrum, 1998a). This can be fulfilled with the consumption of about 327 kg of wheat per year. Calculations based on safe levels of food intake focus on this basic food function. Food security studies have assessed these levels from an agricultural point of view (e.g. Penning de Vries et al., 1995), in which diets were simplified. This diet, however, can only be maintained for a short period of time because it lacks many essential nutrients. Food requirements on the second level, the subsistence level, are based on a selected number of nutrient-dense foods such as milk products, meat, and vegetables. It is optimal from a nutritional point of view and provides health for the total life span (Voedingscentrum, 1998a). 1.4.2 Food consumption patterns This thesis defines food consumption patterns as repeated arrangements observed in food consumption by a population group. They are embedded in types and quantities of foods and their combinations into different dishes or meals. Food consumption patterns depend on several factors such as personal preference, habit, availability, economy, convenience, social relations, ethnic heritage, religion, tradition, culture, and nutritional requirements (De Wijn and Weits, 1971; Ivens et al., 1992; Whitney and Rolfes, 1999; Van der Boom-Binkhorst et al., 1997; Vringer and Blok, 1995; Von Braun and Paulino, 1990; Musaiger, 1989; Wandel, 1988; Von Braun, 1988). Until recently, consumption patterns were strongly influenced by the local availability of commodities, resulting in large regional and inter-generational differences (Jobse-van Putten, 1995). During the 20th century, modern transportation and food conservation techniques resulted in more varied consumption patterns. When health issues are taken into account, and social and cultural functions of food have also to be fulfilled, a menu of wheat or a diet based on a selected number of nutrient-dense foods does not suffice. The actual food patterns are much more varied. Diets on the third scale level (the cultural level) contain foods that are low in nutrient-density such as coffee, cakes, or chocolate,. They also contain larger amounts of foods than diets on the subsistence level. Due to this cultural element, consumption patterns differ strongly among communities and generations. Even among affluent countries, such as the countries of the European Union, large differences exist in the consumption of specific foods (LEI-DLO/CBS, 1998; FAO, 1999). For example, in 1995, the consumption of coffee was 9.4 kg per capita in Sweden, whereas in Ireland, only 1.8 kg was consumed. 11..55 ►► FFoooodd aanndd ssuussttaaiinnaabbiilliittyy Food production results in environmental impacts ranging from local and regional, to global scale levels. Impacts include pollution, such as pesticide and herbicide emissions, malodor, reduces biodiversity, and contributes to climate change through emissions of methane, dinitrogen oxide, and carbon dioxide. Moreover, it puts large claims on finite resources, such as agricultural land, fresh water, and fossil energy carriers. These negative impacts make food production unsustainable from an environmental perspective. However, people need to eat every day. Insight into the impact of food consumption patterns on the environment might provide a tool to change these patterns into a more sustainable direction, with efficient use of natural resources and equitable social development. This thesis examines sustainability from the perspective of consumers. In this way, it fits in the line of research activities of the Center for Energy and Environmental Studies (IVEM) of the University of Groningen in the Netherlands, where much work has been done to investigate the relationship between energy requirements and consumption patterns. The thesis of Kramer (2000), for example, has made an analysis of energy requirements for Dutch food consumption patterns. This thesis continues the IVEM line of research. It extends the research area from energy to two natural resources, land and fresh water, and includes food consumption patterns from all over the world.

Chapter 1 __________________________________________________________________________________________

4

11..66 ►► SSccooppee ooff tthhee tthheessiiss Food contributes to an important share in the total use of natural resources, such as land (e.g. FAO, 2002), fresh water (e.g. FAO, 2003Aa; Falkenmark, 1989b; Rosegrant and Ringler, 1998; Rockstrom, 1999), and fossil energy carriers (e.g. Kramer, 2000). Since the world population is expected to increase and, for many people, food consumption patterns are likely to move in more affluent directions, the pressure on finite natural resources will increase. The adoption of the definition of sustainable development as ‘development that meets the needs of the present without compromising the ability of future generations to meet their own needs’ (WCED, 1987) requires that food consumption patterns in developing countries, where hunger and malnutrition are prevalent, move in a direction that meets basic human needs. Since natural resources are finite, for developed countries, this direction of dietary change could mean that their share in the total use of natural resources should decrease in favor of developing countries. This thesis emphasizes food consumption patterns and analyzes the most sustainable directions of change. It requires information on resource use of specific commodities and food items. In combination with information on consumption, it generates insight into the use of natural resources for a specific combination of a consumption pattern and a production system. The comparison of environmental effects of different consumption patterns, however, requires input from one production system, generating relative results. Such an analysis represents several research subjects. The objective of this thesis is threefold:

it aims to design and develop measuring methods for environmental sustainability in food production systems;

it analyzes the effect of food consumption patterns on natural resources; it identifies those food consumption patterns that have the most sustainable environmental

performance under nutritional and cultural constraints. The central research question of this thesis derives from the three subjects referred to above. What are desirable transition pathways towards sustainable food consumption patterns that have the most favorable characteristics in terms of land, fresh water, and energy use while considering nutritional and cultural constraints? The central research question gives rise to several subquestions. Which foods have large contributions to the use of resources? Are there similarities and differences in the use of resources of food categories and among categories? Are there similarities and differences in the use of resources of food consumption patterns and among consumption patterns? What are desirable directions of change for food consumption patterns? To make assessments operational, the thesis had to make choices and limit the number of research subjects. This makes it necessary to make simplifications. Its focus is the development of methodology for assessments and the comparison of food consumption patterns in their impact on natural resource use. Later research can continue in those areas the thesis did not address. In analyzing food systems, it is important to consider (i) the characteristics of the system and the indicators needed for comparison; (ii) the system boundaries; and (iii) the units of calculation. First, it is important to define the characteristics of the system and related indicators the thesis should address. The preceding section showed that environmental impacts of food production systems take place on local, regional, and global levels of scale. Since food more and more concerns global markets and international trade, where finite availability of natural resources is the most explicit characteristic of the system, the selection of the global scale level and the use of the indicators land, fresh water, and energy, provides a good insight from an environmental point of view. Second, the thesis needs to specify the system boundaries. One of the objectives was to identify differences among resource use for food consumption patterns. Since production systems show such large variation one thesis cannot cover, it selected the Dutch food production system of 1990. It derived information from this system because much information is already available, for example for energy requirements, and, because it concerns a system from a western, developed country with high technological standards. Third, the thesis needs to select the units of calculation. It can express system characteristics in terms of actually required natural resources committed to the production of food, such as square meters, or in theoretical units of calculation. As mentioned before, the combination of a specific production and a consumption pattern determines its requirements for resources. Although actual resource requirements of a food system are likely to be most relevant, the system boundary of this

General introduction

__________________________________________________________________________________________

5

thesis, i.e. the selection of the Dutch 1990 production system, makes it necessary to present most results in a relative way, providing theoretical rather than actual requirements. 11..77 ►► SSttrruuccttuurree ooff tthhee tthheessiiss This thesis is structured as follows: Chapter 2 analyses existing measuring methods for environmental sustainability in food production systems. It provides an overview of indicators from the perspective of companies, as well as sustainability issues from the perspective of environmental research. Next, based on a systemic approach, it proposes a measuring method for environmental sustainability that uses three indicators that address global environmental issues: the use of land, fresh water, and energy (from both fossil and renewable resources). Chapter 3 presents a method to determine land requirements for individual food items. It adopts the methodology from earlier energy studies to combine resource requirements per unit of food with household consumption. The chapter applies the method for the Dutch production system of 1990 which results in an overview of land requirements for over one hundred individual food items. Chapter 4 applies the information on land requirements from Chapter 3 to make assessments of theoretical food consumption patterns on the basic and subsistence level, as well as for actual patterns that occur in Europe and the United States. The chapter shows important factors that determine the total land requirement of a food consumption pattern. Chapter 5 develops a method to calculate the growth-related factor of crop water requirements, assesses the impact of crop characteristics on water requirements, and evaluates options to reduce the use of water by changing food consumption patterns. Chapter 6 analyzes the relationship between per capita income, and food supply, the composition of food consumption, as well as the contribution of animal foods. It makes spatial and temporal comparisons for food consumption patterns and identifies patterns of change that have an impact on the need for resources. Chapter 7 integrates knowledge from preceding chapters. First, it compares differences and similarities for land, fresh water, and energy for one specific food consumption pattern and identifies long-term trends. Second, it compares land, fresh water, and energy requirements for a food consumption pattern from a developing and from a developed country. These comparisons provide information on desirable changes towards sustainable food consumption patterns. The chapter, presents a conceptual framework of the food system, discusses the main results and gives suggestions for further research. Next, it presents the main conclusions in answering the central research question of this thesis.

6

7

CChhaapptteerr 22

DDeessiiggnn aanndd ddeevveellooppmmeenntt ooff aa mmeeaassuurriinngg mmeetthhoodd ffoorr eennvviirroonnmmeennttaall

ssuussttaaiinnaabbiilliittyy iinn ffoooodd pprroodduuccttiioonn ssyysstteemmss∗∗

Abstract These days, sustainability is a key issue for many private companies that address their sustainable

corporate performance (SCP). The perspective is essential for their license to operate and forms the basis for

business principles and practices. The lack of internationally accepted reporting standards on what, when, and

where to report makes it difficult to assess sustainability, however. Moreover, measuring tools providing

information on SCP are only the first step towards sustainability. To prevent negative effects of operations being

transferred from one company to another, the second step is the development of a system-based approach for all

companies that contribute to an end product. This chapter presents the findings of a study about the use of

environmental indicators for food production, and proposes a measuring method for environmental sustainability

in food systems. The chapter shows that environmental SCP often focuses on events at a local level. The

enormous number of indicators found in literature generates too much data that often provide no additional

knowledge on environmental sustainability of a system. Moreover, although environmental research has

addressed many aspects of sustainability, it has often ignored interactions. Overall environmental implications of

food production are therefore poorly understood. The proposed measuring method uses three indicators that

address global environmental issues: the use of energy (from both fossil and renewable sources), land, and water.

The systemic approach can calculate trade-offs along supply chains that make up a production system. The use

of the method implies an extension of environmental SCP towards the overall performance of a production

system. The final outcome is expressed in three performance indicators: the total land, energy and water

requirement per kilogram of available food. For companies, the data generated can be used to compare trends

over time, to compare results with targets, and to benchmark a company against others. For consumers, data can

be used to compare environmental effects of various foods. The method is also applicable for other business

sectors. Results presented in this chapter are part of a multidisciplinary project on the scientific modeling and

measuring of SCP involving economic, social and environmental dimensions. Acceptance of the measuring

methods developed may be a powerful contribution towards creating sustainable business practices.

22..11 ►► IInnttrroodduuccttiioonn During the last decades, environmental issues have evolved from pollution and depletion of natural resources towards global issues, such as climate change. Three important milestones have been: the identification of chlorinated pesticides as major pollutants in ‘Silent spring’ (Carson, 1962), the notion that non-renewable natural resources can become depleted (Meadows, 1972), and the introduction of the sustainability concept in the ‘Brundtland report’ (World Commission on Environment and Development, 1987). ∗ This chapter is a slightly adapted version of Gerbens-Leenes, P.W., Moll, H.C., Schoot Uiterkamp, A.J.M., 2003. Design and development of a measuring method for environmental sustainability in food production systems. Ecological Economics 46, 231-248.

Chapter 2 _________________________________________________________________________________

8

Private companies have attempted to respond to environmental issues. Nowadays, the economic performance of business is often seen in conjunction with its social and environmental performances (Steg et al., 2001). Environmental sustainable corporate performance (SCP), defined as ‘Good housekeeping’ through prevention of pollution and waste and efficient use of scarce resources, is considered important for every company. For many companies, this new perspective is essential for their license to operate and forms the basis for business principles and practices. The perspective meets societal demands for responsible business behavior, and it fits into a modern culture requesting accountability and transparency from powerful institutions. The increasing influence of companies on societies all over the world should go along with their own increasing responsibility and accountability (SER, 2001). In some countries, however, regulations are sometimes less developed than the standards of companies. Therefore, proactive management is sometimes needed because regulatory compliance is not always sufficient to manage the negative environmental or social impacts of business operations effectively. Failure to manage these impacts raises three serious risks: the threat of increased regulatory control by national governments and international organizations; financial risks caused by pollution and large resource use; and damage to the corporate image (Rondinelli and Berry, 2000). Important developments for the issue of sustainability were the foundation of the World Business Council for Sustainable Development (1997), the foundation of the Global Reporting Initiative (GRI, 2000), and the development of standards for environmental management systems, such as the ISO and EMAS standards (OECD, 2001). However, the lack of internationally accepted reporting standards on what, when, and where to report makes it difficult to assess sustainability. For the manufacture of a final consumption item, many processes take place in several companies that form a production system. These processes generate a large variety of impacts. Existing sustainable business practices tend to focus on company performance rather than on performance of the production system as a whole. This has two important disadvantages. First, if companies center on impacts generated by their own activities, large company efforts may still result in small improvements in the production system. Second, the focus on company performance often implies that impacts on a local level of scale, for which the company is responsible, are addressed. Impacts on a global level of scale, for which all companies in a production system are responsible, require a systems approach resulting in a shared responsibility. An important step towards sustainable business practices is the design and development of a system-based measuring tool providing information on the sustainability of all companies that contribute to an end product. A technique for assessing the potential environmental impacts associated with the manufacture of a product is life cycle assessment (LCA) (Heijungs et al., 1992). The systems approach considers all the companies involved, follows products, materials, and substances from cradle to grave, and assesses relevant physical flows (Moll, 1993). The approach prevents that the negative effects of operations being transferred from one company to another or to consumers. A systems approach implies that companies not only focus on their own performance, but also state what they expect of business partners. For example, some companies demand that their suppliers refrain from child labor or create good working conditions (OECD, 2001). While companies have often focused on their own performance, sustainability related scientific research has mainly addressed isolated issues such as, for example, climate change and related fossil energy use. Scientific research should also address sustainability issues in an integrated way, identify interplay and avoid ‘problem shifting’. At present, it is not possible to measure sustainable, corporate performance from a system perspective because an integrated measuring method addressing the three aspects of sustainability - economic, social and environmental - has not yet been designed and developed. There are many different and often rather complex production systems manufacturing an enormous variety of consumer items. This chapter focuses on one complex system, the production of food. Food production requires the input of natural capital, such as land and water, but also of energy provided by the natural capital. A doubling of global food demand is expected in the next 50 years. This poses huge challenges for the sustainability of food production (Tilman et al., 2002). Environmental impacts of agriculture often remain unquantified and therefore do not influence farmer or societal decision-making about production methods. This chapter presents the findings of a study about the use of environmental indicators for food production. It proposes a measuring method based on a small set of system-based indicators for the assessment of sustainable, environmental performance in food production systems. The study was part of a multidisciplinary research project on the definition and measuring of SCP. The project defined sustainability in three dimensions: economic, social, and environmental. It focused on the definition of SCP, and on the development of a practical measuring system. SCP was defined in relation to the potential addition of economic, social and environmental value to the society through corporate activities. These three ‘added values’ are the components of the ‘sustainability value added’

Design and development of a measuring method for environmental sustainability in food production systems

_________________________________________________________________________________

9

of a company. The general findings of the SCP project are reported elsewhere (e.g. Steg et al., 2001). The specific aims of this chapter are:

to identify and analyze methods used so far to describe and improve environmentally sustainable business practices for food production;

to identify the key global issues relevant for food production from the perspective of the main functions of the natural capital needed for the proper function of society;

to design and develop a measurement method based on a small set of system indicators that address the main environmental impacts of food production.

The measuring method proposed here considers the entire production system and addresses global environmental issues. It calculates the use of resources for food production, but it is also applicable for other production systems. Acceptance of the measuring methods developed in this chapter may make a powerful contribution towards creating sustainable business practices. 22..22 ►► SSyysstteemm ddeessccrriippttiioonn Food constitutes an important and indispensable group of consumer items. In the Netherlands in 2002, for example, households spent 17% of their budgets in this category (CBS, 2002). Food items are manufactured in a complex system made up of many processes in several supply chains. To manufacture a cake, for example, requires various commodities from agriculture, including wheat, sugar beet, milk, and eggs. The food industry processes some commodities to produce basic ingredients for cakes: sugar, flour, and butter. Finally, using ingredients from different supply chains, bakeries make cakes. Transportation by airplane, boat, train, or truck provides the global availability of commodities and foods. The food system encompasses production, consumption, and final waste handling. Figure 2.1 shows a raw materials production chain of the system, the system boundary of this study, and the network of supporting business sectors. The purpose of this chapter was to design and develop a measuring method for environmental sustainability in food production systems. Consumption, waste handling, and the supporting network fall beyond the scope of the study because the application of the method is targeted to company performance. The food production system consists of a huge number of processes. A process is defined here as an activity with an identifiable beginning and end that takes place at a certain location, with a fixed ratio of input and output flows. The output of one process is the input of the next one. Processes can sometimes be broken down into discrete sub-processes or unit operations as defined in the chemical process industry. The baking of a cake, for example, consists of several sub-processes, such as mixing the ingredients and putting the cakes into an oven. This chapter distinguished three scale levels for the food production system: the first level is the raw materials process level, the second level is the raw materials chain, and the third level is the food production web. At level one, processes for the manufacture of raw materials take place. At level two, a series of processes form a linear production chain and provide raw materials. In contrast to processes, production chains have variable input and output flows. At level three, raw materials originating from several chains join and form a complex production web. Physical streams sometimes flow in the opposite direction. At this level, the process of the manufacture of a final food item requiring input from more than one chain takes place. Figure 2.1 shows a raw materials production chain at level two and two return flows at level three. Private companies are defined here as business units in which production and transportation processes take place that contribute to the manufacture and availability of a final food item. Some companies are responsible for only one process while others control complete production webs. The complexity of the food system is demonstrated by an example, the manufacture of a Dutch cake. Processes take place in several companies in various sectors: a French farm grows wheat, a Dutch flour manufacturing company processes the wheat, a bakery manufactures the cakes, and several transportation companies bring raw materials from supplier to producer. At level one, processes and sub-processes contributing to the manufacture of raw materials, like sowing wheat and processing sugar beet, take place. At level two, the production of flour is an example of a raw materials production chain. First, the wheat is grown. Second, the wheat is transported to the Netherlands. Third, the wheat is processed to make flour. At level three, the web level, the chains necessary for the raw materials for cakes are joined, such as production chains for eggs, sugar, milk, and butter. At this level, the process of baking the cake, the final food item, takes place. Sometimes physical flows in the opposite direction occur; for example, waste streams from the sugar industry can be reused for livestock fodder. Fed to cows, these waste streams contribute to butter manufacture for cakes.

Chapter 2 _________________________________________________________________________________

10

Fig. 2.1. Overview of the food production system, the system boundary, the output to consumers and the network of supporting business sectors that also provide their services to other production systems. Production processes take place in several business sectors represented by the boxes. A series of processes forms a production chain. The arrows show transportation of physical streams between the links of the chain. Closed arrows represent transportation in one direction at level two, open arrows represent physical streams in the opposite direction at level three. In a production web, all chain links and transportation activities between links contribute to the overall environmental impact of a food item. For the production of a package of frozen French beans at level two, for example, vegetables are grown, processed, packed, transported and stored. These processes require the input of scarce natural resources, such as land, energy carriers, and water, and contribute to pollution (Tilman et al., 2002). The environmental impacts of processes in chain links or of transportation between links often differ considerably. This is illustrated by an example taken from energy studies (Kok et al., 2001). Table 2.1 shows the energy requirements for processes and transportation in production chains of Dutch vegetables. Table 2.1 reveals large differences among the specific energy requirements for production processes and for transportation modes (MJ per kg, or MJ per 1000 kg per km). The final energy requirement of 1 kg of vegetables varies by a factor of 15. Fresh vegetables produced in Africa and transported by airplane to the Netherlands require 88 MJ per kg, whereas locally produced open air vegetables require only 6 MJ per kg (Kramer et al., 1994). In 1990, an average Dutch household consumed 162 kg of vegetables with a related energy requirement of 2500 MJ (Gerbens-Leenes, 1999). The example shows the necessity of a system-based approach for the assessment of sustainability related to food production. Not only is the performance of an individual company or business sector important but also the overall performance of all companies in a production chain or web. Differences among production methods, transportation modes, and distances heavily influence the environmental pressure of a final food.

Food Production Chain

Crop Production

Livestock Production

Food Processing Industry

Trade and Retailing

Consumer

Waste Handling

T

T

T

T

Primary Extraction Industry

Power Generation

Banking, Marketing and

Insurance Industry

Supporting Network

System Boundary

= Transportation T


_________________________________________________________________________________

11

Table 2.1 Overview of energy requirements in production chains of vegetables available in the Netherlands (Source: Kok et al., 2001)

Process Energy requirement (MJ per kg vegetables)

Agriculture

Open air production 0.7 Greenhouse production 26.2

Transportation Energy requirement (MJ per 1000 kg/ km)

Ship (inland shipping) 0.6 Ship (sea, bulk) 0.1 Lorry 1.7 Train 0.6 Airplane 9.0

22..33 ►► MMeetthhooddss 2.3.1 Analysis of existing measuring methods and scientific research issues Business mainly addresses sustainability issues using a bottom-up approach and focuses on the performance of individual companies. Scientific research, on the other hand, often focuses on sustainability issues on a global level using a top-down approach. This chapter inventoried and analyzed these two approaches. It performed a literature search of scientific publications on items concerning sustainability from the perspective of sectors involved in food production. These sectors were agriculture, transportation, manufacturing, and retailing. Environmental status reports of individual companies fell beyond the scope of this thesis. A recent OECD report (2001) provided additional information. The results give an impression of the efforts made so far to measure and report on sustainability issues for food production. This chapter identified existing environmental sustainability indicators and clustered them according to their level of scale. These levels were the local, regional and global level. The analysis of the top-down approaches focused on the identification of global issues relevant for food production identified by environmental researchers. The section ‘Results’ presents the findings of the literature search. 2.3.2 Design and development of an environmental measuring method To design and develop a system-based measuring method for food production, this chapter used a combination of top-down and bottom-up approaches. The starting point was the selection of a small set of environmental sustainability indicators addressing issues on a global level relevant for food production. The bottom-up approach implies that actual measuring takes place in companies. The measuring method provides information on three system levels: the process level, the raw materials production chain level, and the production web level. For a company, information about environmental performance becomes available for benchmarking, or for the improvement of performance over time. At the production chain level, information about the environmental pressure of suppliers becomes available. At the web level, the combination of the performances of all companies reveals the total environmental pressure related to the production of a final food. The section ‘Design and development’ presents a detailed description of the method. 22..44 ►► RReessuullttss 2.4.1 Environmentally sustainable business practices The literature search revealed a large number of publications on the relationship between sustainability and business. The following presents the main results that apply to all company types. Some studies have identified consumer demand for quality products that includes environmental requirements (Stauffer, 1997; Boudouropoulos and Arvanitoyannis, 1999). Societal demand has led

Chapter 2 _________________________________________________________________________________

12

companies to recognize that proactive environmental management leads to profitable results (Stauffer, 1997). A study on environmental supply chain dynamics (Hall, 2000) has shown that environmental change within a supply chain can be stimulated by a so-called ‘channel leader’ with sufficient power over suppliers. These ‘channel leaders’ must be under specific environmental pressure themselves. A number of studies have found that one of the most significant pressures forcing firms into addressing environmental concerns was the emergence of the ‘green consumer’ ( Williams et al., 1993; Steger, 1993; Drumwright, 1994; Elkington, 1994). Consumer pressure, however, is almost entirely focused on recognizable consumer goods, often associated with large multinationals (Hall, 2000). Firms that operate far from the end consumer and are hidden within a supply chain are under little environmental pressure. For example, it is unlikely that airline passengers fully understand the environmental impacts of aircraft manufacturing, and would therefore exert little pressure on airlines to purchase ‘green aircraft’. Other studies have demonstrated that the economic, social, and environmental impacts of multinationals cause concern (OECD, 2001). Many firms have responded to these concerns with managerial innovations, including codes of conduct. This chapter found three important steps towards the measuring and reporting of sustainability. The first step was in the 1970s. Companies started to issue policy statements or principles (OECD, 2001). These principles are codes of conduct, stating commitments on business ethics and legal compliance. The first corporate code of conduct was the 1977 ‘Issuance of guidelines on conducting business in South Africa’ by an automobile manufacturer. Later, many other companies adopted these ‘Sullivan Principles’, or began to issue corporate codes dealing with business ethics. The second step was the development of management systems or practices that refer to action strategies and programs. More recently, the third step formulated the outcomes, standards providing guidance for business reporting on non-financial performance. The literature search showed, however, that many companies mainly focus on their own performance, and that only some firms feel responsible for their suppliers’ activities (Hall, 2000). At the end of the 20th century, many multinationals certified their environmental management systems (EMS) under ISO 14000 standards, and many others were in the process of doing so (Rondinelli and Vastag, 2000). Nowadays, an increasing number of companies publish information on environmental impacts of their activities, the outcomes. According to the OECD study (2001), however, the absence of internationally agreed reporting standards on sustainability results in a range from rudimentary reporting to full-scale reporting. For example, only 17% of European companies and 41% of European high environment impact (HEI) companies reported in some way on their environmental performance. Moreover, companies made their own choices regarding the scope and depth of reporting. Of all the companies that reported on environmental performance, 62% provided some quantitative data while only 15% of these companies reported on all key issues. 2.4.2 Sustainability indicators from the perspective of companies The chapter found many attempts to develop tools to measure sustainable business performance. This is in line with the OECD report (2001). For example, the International Standards Organisation (ISO) ; (Boudouropoulos and Arvanitoyannis, 1999; ISO World, 2000), the European Union’s Eco-Auditing Management System (EMAS) (Kolk, 2000; OECD, 2001), and the Lowell Center for Sustainable Production (LCSP) (Veleva and Ellenbecker, 2001) have developed indicators for sustainable production. The LCSP has presented 22 indicators on five levels of scale that can be calculated as totals or per unit of a product. To operationalize business principles and to measure performance, business sectors and scientists have developed many environmental indicators. Table 2.2 presents an overview of frequently used indicators for environmental sustainability in food production. First, Table 2.2 shows indicators according to the business sector they apply to. Second, it ranges indicators according to three levels of scale: the local level, the regional level, and the global level. For the agricultural sector, Table 2.2 shows indicators proposed by agricultural researchers, the Dutch government, and the European Union (EU). Many of these indicators address issues at a local or regional level, such as pollution and the quality of resources. This results in a large number of indicators. At a global level, indicators address the use of phosphorus, land, energy, and emission of greenhouse gasses. Although sustainability concepts exist for agriculture, they do not necessarily imply a measurable set of indicators for their characterization. Some concepts are philosophical by their very nature (Hansen, 1996), and therefore difficult to measure. In his overview article, Hansen (1996) described these philosophical approaches to sustainable agriculture. Approaches are often contrasted with conventional agriculture that is characterized as capital-intensive, large-scale, with extensive use of artificial fertilizers, herbicides, and pesticides, and intensive animal husbandry.


_________________________________________________________________________________

13

Environmental values associated with sustainability include mimicry of nature and an ecocentric ethic. That overview has also shown that some other concepts interpret sustainability as a set of strategies providing useful, measurable indicators. For the transportation sector, indicators have not only been proposed by the EU, governments, and environmental researchers but also by business researchers. Table 2.2 shows that most indicators address pollution and the related quality of the urban environment on a local level, whereas relatively few indicators address issues at regional or global levels. For the manufacturing industry, business researchers, among others, have described the relationship between companies and the environment. Here, indicators mainly address emissions by the company on a local level. Furthermore, energy use and climate change have been recognized as environmental problems. In order to control performance, some companies developed environmental performance indicators and recognized the importance of the environmental performance of suppliers (Thoresen, 1999). The EU in particular recognizes many environmental issues related to the manufacturing industry. For the retailing sector, environmental issues in preceding links in the food chain are very important (Bansal and Kilbourne, 2001). Retailers pay less attention to environmental impacts related to their own activities, such as distribution, choice of location, and merchandising, and pay more attention to environmental effects in preceding links in the supply chains, such as agriculture. Table 2.2 shows that further down a production chain, the environmental performance of the whole chain becomes more important than the performance of an individual company. In the first chain link, agriculture is mainly concerned with the environmental effects of their own performance. Transportation also focuses on own performance. Although the manufacturing industry mainly focuses on own performance, suppliers’ performance is also important. In the last chain link, retailing is mainly concerned with the environmental effects of upstream activities. 2.4.3 Sustainability issues from the perspective of environmental research Environmental research recognizes the importance of chain management for the sustainability of food production. Life Cycle Assessment (LCA), for example, is a technique for assessing the potential impacts associated with a product, by compiling an inventory of relevant environmental exchanges of the product throughout its life cycle (‘cradle to grave’) and evaluating the potential environmental impacts associated with those exchanges (Heijungs et al., 1992; Weidema, 1999; Weidema and Meeusen, 2000). It is argued that in order to make food systems more sustainable, it is important to avoid trade-offs between chain links. Many studies have addressed energy and some have developed concepts like indirect energy use (e.g. Wilting, 1996; Wilting et al., 1999; Carlsson-Kanyama and Faist, 2000; Kok et al., 2001). In these concepts, all energy use in and between chain links contributes to the final energy intensity (MJ per financial unit), or the final energy requirement (MJ per physical unit) of a product. These studies formed the basis for a large number of studies addressing energy reduction strategies in food chains (e.g. Kok and Kramer, 1995; Carlsson-Kanyama, 1997, 1998; Gerbens-Leenes, 1999; Andersson, 2000; Dutilh and Kramer, 2000; Gerbens-Leenes, 2000). Several studies have recognized the importance of land use for food production. For example, Wackernagel et al. (1997) have developed the concept of the ecological footprint that evaluates human land use. Some authors have shown the importance of food consumption patterns on land requirements (Gerbens-Leenes, 1999; Van Vuuren and Smeets, 2000; Gerbens-Leenes et al., 2002), while agricultural studies have demonstrated the relationship between food security and land use (Penning de Vries et al., 1995; Bouma et al., 1998a; Groot et al., 1998). Andersson (2000) has proposed the application of LCA to food production systems, and has recognized the importance of energy, land, water, and nutrients for the sustainability of these systems. 22..55 ►► DDeessiiggnn aanndd ddeevveellooppmmeenntt ooff aa mmeeaassuurriinngg mmeetthhoodd The measuring method proposed here was part of a multidisciplinary project on the scientific modeling and measuring of SCP aimed at the development of a three-dimensional model involving economic, social, and environmental dimensions. Some environmental effects of individual companies become manifest at local and regional scales. Examples are noise, malodor, and emissions of locally polluting substances. The overall project considered these effects to be indicators of social SCP, and addressed them in the social part of the SCP project. This chapter focused on the integral environmental effects of operations of all companies that contribute to the manufacturing of a final food on a global scale leading to the concept of shared responsibility. It designed and developed a measuring method in three steps. First, this chapter integrated bottom-up and top-down approaches,

Chapter 2 _________________________________________________________________________________

14

second, it selected relevant indicators, and third, it designed a flow chart of calculations and inputs thereby adopting the life cycle approach and allocation methodology from LCA. 2.5.1 Integration of bottom-up and top-down approaches Table 2.2 demonstrates that environmental corporate performance mainly focuses on local events, such as pollution by emissions. This leads to an enormous variety of indicators that are often not in line with general constraints for indicators, that focus on company rather than production system performance, and that generate too much data. As a result, too much information is generated that in many cases provides no additional knowledge on the environmental sustainability of the production system as a whole. Therefore, the environmental effects on the web level, to which all companies contribute, as well as global issues, are paid too little attention. Environmental research has addressed many aspects of sustainability, such as energy use and global warming. It has calculated the trade-offs between chain links for the environmental effects of production, but it has often ignored the interactions between aspects of sustainability. When calculating the energy reductions for food, for example, environmental studies have not considered the effects of these strategies on other resources (e.g. Kramer, 2000). Overall environmental implications of food production, therefore, are poorly understood. A measuring method for environmental sustainability in food production systems should also addresses these interactions. LCA has already developed methods to evaluate the environmental impacts related to the manufacturing of products. The basis for the calculations is the life cycle approach. Environmental effects caused by processes that generate more than one output are assessed by the method of allocation (Heijungs et al., 1992). In general, LCA is independent of the location of a production system and assumes a linear relationship between the amount of product manufactured and the environmental impacts (Wegener Sleeswijk et al., 1996). This chapter proposes a measuring method that integrates top-down and bottom-up approaches, using the strengths of both. It adopts the methodology of the life cycle approach and allocation from LCA. First, it identifies main environmental issues on a global level. Second, it proposes a small set of indicators to measure these issues. Third, it upscales measurements in companies for individual processes to the total production system. 2.5.2 Selection of indicators Natural capital is a key concept in ecological economics (Costanza and Daly, 1992) and refers to the possibility of the natural environment to provide products and to perform functions essential for human existence (Ekins et al., 2003). There are four types of natural capital, air, water, land, and habitats (Ekins and Simon, 2003) that perform four important functions for society. These are the ‘source function’, the ‘sink function’ (Daly, 1990), the ‘life support function’ (Van Dieren, 1995; Ekins and Simon, 2003), and the ‘human health and welfare function’ (Ekins and Simon, 2003). The ‘source function’ refers to the delivery of natural resources to the economy, such as energy carriers, agricultural land, or biological resources. The ‘sink function’ implies the possibility to dispose of waste. The ‘life support function’ addresses a set of functions performed by land, water, and air essential to sustain life. The ‘human health and welfare function’ refers to services which maintain health and contribute to human well-being. According to Daly (1990), the sustainable use of natural resources has three implications: (i) renewable resources should not be exploited at a greater rate than their regeneration level; (ii) non-renewable resources should not be depleted at a greater rate than the development rate of renewable substitutes; and (iii) the absorption and regeneration capacity of the natural environment should not be exceeded. The four types of natural capital are all essential for the production of food, while the four functions are all important for food production and consumption. For the selection of sustainability indicators for the food production system, this chapter used three criteria: general functions and constraints of indicators, essential natural resources for food production, and the functions of the natural capital to society. For the selection of indicators, the chapter used four constraints: (i) indicators provide relevant information about the sustainability of the system, (ii) reliable and accurate measurement is possible, (iii) data are available, and (iv) information can change management choices and optimize production. For the selection of essential resources, the thesis addressed three important requirements for food production: agricultural land, fresh water, and energy input. For land and water input, the human use of these resources is mainly dominated by food

Design and development of a measuring method for environmental sustainability in food production systems _____________________________________________________________________________________________________________________________

Table 2.2 Overview of frequently used indicators for environmental sustainability in food production systems Local Level Regional Level Global Level

Agriculture Pollution: emission of NH3, NOx (I: a, b, II, III); SO2 (II); pesticides (I: a, b, c, II); N, P (I: a, b, c, II); heavy metals (a, c) Efficiency: N efficiency (I: c) Depletion of resources: concentration soil organic matter (I: b); alteration of ion balance (I: a); salinization (I: a); alteration of biological cycles (I: a); soil cover (I: b); % weeds in grain crops (c)

Depletion of resources: ecological structures (I: b); % unsprayed area (I: c); % uncultivated area (I: c); land use (I: e, f); water use (II); crop diversity (I: b)

Depletion of resources: phosphorus use (I: b); land use (I: e, f); energy use (I: b, c, II) Climate change: emission of methane (I: d, III); N2O (I: d); CO2 (I: b, d, II, III)

Transportation Quality of urban environment: congestion, deterioration of streets, public places and architectural heritage (a, b); noise (a, b, c, d); malodor (c) Pollution: emission of: lead (a); NOx; volatile organic compounds (VOC), particulates (a, b, c, d); CO (a, c, d); hydrocarbons (HC) (a, b, d); SO2 (a, c, d); mean values of ozone (a, d); soil pollution by petroleum product disposal, sulphuric acid leaks, heavy metal sludges (d)

Depletion of resources: land use (a, b, c) Pollution: oil spills, loss of cargo, operational pollution from shipping activities (a)

Depletion of resources: material use (c); natural resource degradation (d) Climate change: burning of fossil fuels and related emission of CO2 (a, b, c, d); emission of N2O, CH4, HFK’s, PFK’s, SF6 (c)

Manufacturing industry Waste treatment: reuse, recycling, incineration, deposition (I, a) Environmental management system performance: performance total system, organization of product development/design Production planning: rejects and material waste performance (I) Production: environmental performance of processes, number and consequences of environmental incidents (I, a) Pollution: emission of VOC’s, dioxins, heavy metals, SO2, NO2, particulates, O3, CO, development of cleaner technologies (a)

Procurement planning: supplier selection, surveillance of supplier’s performance (I) Transport planning: transport media selection, environmental performance of transport media in use (I) Pollution: emission of SO2, water quality (a) Depletion of resources: protection of habitats, corridors, and endangered species; water use (a) Other issues: reduction of animal experimentation (a)

Energy planning: total energy consumption, energy mix, consumption of non-renewable energy sources, energy efficiency (I)

Retailing Pollution: pesticide and herbicide emissions, water quality (a, b, c, d) Quality of urban environment: noise and traffic congestion, aesthetics (a)

Depletion of resources: reduction of biodiversity, water depletion (a); forest depletion (a, b, c, d) Other issues: animal welfare, biotechnology (a, b, c, d); inefficient agriculture (a)

Pollution: ozone depletion (a) Depletion of resources and climate change: oil and gas depletion, use of fossil fuels (a)

The table shows indicators according to the business sector they apply to: agriculture, transportation, manufacturing industry and retailing. Indicators are ranged according to three levels of scale: the local level, the regional level and the global level. Agriculture: I. Scientists, e.g. a. Giupponi (1998); b. Bockstaller et al. (1997); c. Halberg (1999); d. (Kramer et al., 1999); e. Gerbens-Leenes (1999); Wackernagel et al. (1997). II. The Netherlands government (Source: Nationaal Milieubeleidsplan 3, (Boer et al., 1998)). III. The EU (Source: EUR-Lex (1993)). Transportation: a. EU (Source: (European Community, 1993)); b. the Netherlands government (Source: (Netelenbos et al., 1999)); c. environmental scientists (Source: (Bouwman, 2000)); d. business scientists (Source: (Rondinelli and Berry, 2000). Manufacturing industry: I Source: Thoresen (1999); a. EU (Source: (European Community, 1993). Retailing: a. Bansal and Kilbourne (2001); b. (Drumwright, 1994); c. (Steger, 1993); d. (Williams et al., 1993).

Chapter 2 __________________________________________________________________________________________

16

scarce resource. It is estimated that the global agricultural land area is decreasing at a rate of 7 % per decade (Oldeman et al., 1991). Land is needed for food but, in the near future, probably also for the generation of renewable energy. The area to be set aside for renewable energy, such as energy generated by wind, biomass, photovoltaic cells, and water, is substantial. The chapter therefore proposes the quantity of land use as an indicator for the use of scarce natural resources of the ‘source function’, as an indicator for the ‘life support function’, and as an indicator for the potential competition among agricultural interests, energy production capacity, and the requirement of natural areas for biodiversity. The second essential requirement for food production is the availability of sufficient fresh water of adequate quality. In large parts of the world, interaction among plants, land, and water sometimes causes irreversible environmental effects that threaten food production (Falkenmark, 1989a). In some parts of the world, water is a scarce resource in a highly competitive context. Water for urban supply or for industrial activities then competes with the large quantities needed for agriculture. In many countries, there is already a scramble for land (Falkenmark, 1989a). If a ‘water scramble’ is not to follow, extremely wise consideration of technical, financial, economic, political, legal, and administrative issues is needed. This chapter, therefore, proposes the quantity of fresh water use as an indicator for the use of scarce, high quality, fresh water resources of the ‘source function’ and of the ‘life support function’. The third essential requirement for food production is energy, generated either by fossil energy carriers or by renewables, such as biomass. The consumption of food has a considerable impact on household energy use (Kramer, 2000). An average EU household, for example, requires 19% of its energy use for food (Nonhebel and Moll, 2001). This chapter, therefore, proposes the quantity of fossil energy use as an indicator for the depletion of non-renewable resources of the ‘source function’ (IPCC, 1996), and for possibilities of the ‘sink function’ to neutralize climate change and acidification. The quantity of renewable energy use is also an indicator for the ‘life support function’, and for the potential competition among agricultural interests, energy production capacity, and the requirement of natural areas for biodiversity. An important criterion for the selection of the indicators land, energy, and water for the measuring method for environmental sustainability in food production systems is the interaction among these indicators. For example, ‘high external input farming’ (HEI) requires large energy input in the form of fertilizer and pesticides to reach high production levels. As a result, land use is relatively low. On the one hand, irrigation in agriculture results in higher yields and thus lower land use, but on the other hand, in larger energy and water requirements. The use of the three indicators proposed in this chapter also addresses interactions among aspects of environmental sustainability. 2.5.3 Flow chart of calculations and input For the assessment of the environmental sustainability of food production, a step-by-step approach, in which the output of one step forms the input of the next one, calculates total environmental system performance. Figure 2.2 shows an overview of the three levels of scale, three calculation steps, and the required input. At level one, processes take place in companies; at level two, processes for the manufacture of one raw material take place in chains; at level three, production chains join and form a web in which a final food item is produced. The outcome of the combination of direct company input with indirect chain and web input is the total land, energy, and water requirement per kilogram of available food. The total energy, land, and water requirements for a specific food item are calculated in three steps as follows: Step one: n n DRx = ∑ Ix,s * EREx,s

(x = energy) or DRx = ∑ Ix,s (x = land or water) s=1 s=1 where DRx is the total direct input per indicator x per company (MJ, m2, m3) per year; Ix,s is the direct input per indicator x per proces s (MJ, m2, m3) per year; and ERE is Energy Required for Energy is amount of primary energy per proces s needed for the production of the energy used (MJ per MJ).


__________________________________________________________________________________________

17

Step two: n TRx = ∑ Rx,i * Ai + DRx i=1 where TRx is the total input per indicator x per company (MJ, m2, m3) per year; Rx,i is the specific, indirect requirement for indicator x per kg of physical input i (MJ, m2, m3) per kg per year; and Ai is the total amount of input i needed for the production (kg per year). Step three: Eo / Et * TRx Ro = ------------------ Oo where Ro is the specific requirement for indicator x per kg of physical output o (MJ, m2, m3) per kg; Eo is the economic value of output o; Et is the economic value of the total output; and Oo is the amount of output o (kg per year). Fig. 2.2. Overview of the three levels of scale, three calculation steps and required input for the calculation of the environmental sustainability of a food item. Level one is the raw materials process level; level two is the raw materials chain level; and level three is the food production web level. Step one calculates direct resource use per company, step two calculates resource use per production chain and step three calculates resource use for the final food item. Required inputs are: energy (E), water (W), and land use (L) per company, per chain, and per combination of chains contributing to a final food. The allocation of resource use to food items is carried out according to economic value. Step one calculates direct energy, water, and land use for a specific company per year. Energy use is converted into primary energy use, or in case of renewable energy, into land and water use. Data on basic materials used, and on energy and water inputs are taken from the economic data of the companies concerned. A method to assess land requirements is presented in Gerbens-Leenes et al., (2002). Step two combines direct resource use per company per year with resource use in the preceding chain links. Information on indirect requirements must be obtained from the suppliers in the

Step 1 Level one: Processes

in Companies

Step 2 Level two: Processes

in Chains

Step 3 Level three: Processes

in Webs

Direct Resource use per Company



Use of E, W, L per Process per Year, Energy Required for Energy (ERE

Value)

Direct + Indirect

Resource use per Company

Direct + Indirect

Resource use per Company

Resource use per Food item

Physical Inputs, Indirect Use of E, W, L per Unit of Physical Input

Steps and System levels

Inputs

Physical Output, Economic Output

Chapter 2 __________________________________________________________________________________________

18

preceding links. Step three adopts an allocation methodology from LCA. It assesses resource use per kilogram of a specific food by dividing total direct and indirect resource use over end products according to their economic value. The quantity of land use, water use, and energy use can be calculated for any company along a food production chain. In this respect, attention must be paid to the energy use indicator. A shift in energy use from fossil sources to energy from renewable sources brings with it a requirement for more land, or, in the case of the use of biomass, a need for agricultural land and water. These requirements should also be included in the calculations. Box 2.1 shows an example of the calculation method taken from the energy studies assessing the energy requirement of 1 kg of French beans in a jar (Kramer et al., 1994). It shows how the calculation method works. The example demonstrates that differences in resource use in links of a production chain may have important impacts on the final resource use of a specific food. 22..66 ►► DDiissccuussssiioonn 2.6.1 Present situation: type of information generated and utility The contacts with business (e.g. Royal Ahold), the studies in the economic, social, and environmental fields of the SCP project, and a presentation of results of the project at an international workshop in Groningen in 2001 (Steg et al., 2001) confirmed the general idea that there are large numbers of conceptual frameworks for SCP, but that a generally accepted measuring and reporting system is lacking. Many companies address the sustainability issue, but use an enormous variety of indicators to assess environmental SCP. For three reasons, this generation of large information streams provides hardly any additional knowledge on SCP or on the environmental sustainability of a production system. First, indicators differ among companies and therefore generate incompatible information. Second, companies differ, sometimes even in the same business sector. Outsourcing of specific processes to other companies, for example, is very common. Sustainability, therefore, cannot be assessed at a company level because the type and number of processes performed differ per company. Third, interactions among aspects of sustainability are very important but are often ignored.

Production food industry (FI) 8 MJ

Consumer

Production jar (P) 8 MJ

Retailing (R) 1 MJ

T1 = 2 MJ

T3 = 1 MJ

Indirect input 2 = indirect input 1 + P + FI + T2 + T3 = 22 MJ

+ T2 = 1 MJ

Indirect input 1 = A + T1 = 4 MJ

Output = indirect input 2 + R = total energy requirement of f 1 kg of French beans in a jar = 23 MJ

Calculation of energy

requirements

Production chain

Agriculture (A) 2 MJ

BOX 2.1 Example of the calculation method assessing the energy requirement of 1 kg of French beans in a jar. The output of a production proces is the input of the next one. (Source: Kok and Kramer, 1995; Kramer, 2000)


__________________________________________________________________________________________

19

LCA has developed methods to assess the impacts associated with the manufacture of agricultural products (Weidema, 1999; Weidema and Meeusen, 2000). It uses large indicator sets and provides information on differences of agricultural production systems. LCA assumes a linear relationship between the amount of product manufactured and environmental impacts, and neglects spatial and temporal variation. If only few products are compared, LCA is the appropriate tool. For the assessment of environmental sustainability of complex food production systems, however, it is important to select only few indicators, and take non-linear relationships and natural variation into account. The measuring method proposed here can assess sustainability from a systems perspective because it addresses all the processes in a system in an integrated way and can calculate trade-offs. 2.6.2 The measuring method: type of new information and utility The method presented here is an attempt to measure the environmental sustainability issue in food chains and systems in an integrated way from a global perspective. It is part of a three-dimensional model of SCP involving the social, economic, and environmental dimensions of sustainability. It is stressed that the SCP project considered local environmental effects as indicators of social sustainability and addressed these effects elsewhere. There are three important constraints for a measuring method: it must be simple, it must be easy to use, and it must lead to relevant information on sustainability. The small set of core indicators, land, water, and energy use, is in line with these requirements, and can be calculated for all companies in the food system. The final outcome is the total land, energy, and water requirement per kilogram of available food. Bottom-up approaches have already developed methods to assess environmental SCP on a local level of scale. The indicator set proposed here should complement the indicator sets used so far. In this way, companies can also address sustainability issues on a global level of scale. For companies, the utility of generated information is to compare trends over time, compare results with targets, benchmark a company against others, and assess the relative contribution of the company to a final product. Time trends, for example, provide information on the effects of changes in production methods or the effects of financial investments. On the one hand, environmental pressure is determined by production systems, for example efficiency in the food industry, but on the other hand by food consumption patterns. The influence of consumption on resource use is large. Land requirements for existing European food patterns, for example, differ by a factor of two (Gerbens-Leenes and Nonhebel, 2002). For consumers, the utility of the information is to enable them to compare the environmental effects of various foods and to make well-founded choices. An individual process in a chain link shows little variation over time. For the manufacture of an end product, however, the collective behavior of several processes combined in a production chain can show large variation because very often there are several ways to manufacture a final food item. An important advantage of the systems perspective is that these differences among production chains become visible. The consequence of the use of system indicators is that the sustainability issue is no longer restricted to the performance of a single company but is extended towards the performance of others. This implies a difference between corporate performance and shared corporate responsibility. The former focuses on the relationship between the company and its stakeholders. These stakeholders are often well known, while methods to assess corporate performance have already been developed. Shared responsibility is related to the place of the company in the production system and the responsibility of a company for the performance of other companies contributing to that system. For the sustainability of production systems, there are relatively few stakeholders. Moreover, the sustainability of these systems is difficult to assess and requires other measuring methods.

Chapter 2 __________________________________________________________________________________________

20

2.6.3 Future research: type of additional information required At the moment, it is important to start measurements in companies willing to cooperate, and bridge the existing gap between theoretical scientific knowledge and practical company knowledge. This will take great effort from both sides. In addressing the sustainability issue, research emphasizes the need for accuracy and completeness; business, on the other hand, wants to have an easy to handle, practical, and cheap tool to assess their SCP. The development of such a tool will therefore take some time and adjustments from both sides. At present, there is a relatively small group of companies that are innovative and willing to take the lead in sustainable entrepreneurship (Vollenbroek, 2002). Food companies should realize that proactive, environmental management leads to profitable results. This will become possible when companies start reporting on the set of indicators proposed so that data become available. These data will provide insight on the use of natural capital in specific chain links and will form the basis for optimization strategies. The information generated might lead to differences in reduction strategies because the availability of natural capital shows temporal and spatial variation. This implies that sustainability indicators should not be minimized but rather optimized within environmental constraints. For example, energy reduction strategies in agriculture require larger land inputs. Environmental pressure also differs strongly among business sectors. The emission of wastes, for example, can be a useful indicator for the food processing industry but is less appropriate for the retailing sector. Future research should further elaborate the core indicators proposed in here into sets of supplemental indicators that are only used for relevant business sectors. Conventionally, the three central objectives of sustainable development are environmental, social and economic in nature (World Commission on Environment and Development, 1987). In terms of the contribution of private sector companies to sustainable development, the use of the sustainability concept implies that environmental, social, and economic indicators have to be optimized within existing constraints. Future studies should develop methods that address the trade-offs and interplay between the three aspects of sustainability. However, what is considered important and what is not is a political choice that can be addressed using weighing factors that differ both in time and among regions. Wide acceptance of the environmental measuring method together with methods to measure economic and social dimensions as proposed in the SCP project will make a powerful contribution towards creating sustainable business practices. It can drive towards the decoupling of economic growth and environmental and social degradation. 22..77 ►► CCoonncclluussiioonnss The environmental reporting of companies is still poorly developed. Widely accepted standards for sustainability reports are not available. So far, companies have usually addressed the environmental effects caused by their operations on a local level using a large number of indicators. As a result, the information generated is incompatible and does not address the sustainability issue as a whole. To fully understand the environmental implications of food production, this chapter proposes a measuring method for environmental sustainability using a systems approach. First, it assesses the environmental sustainability of processes in companies. Second, it calculates sustainability for a production system. It reveals the overall environmental effects related to food production, and it expands the SCP of an individual company towards the SCP of all companies contributing to a production system. In this way, companies have a direct responsibility for effects related to their own operations, and a shared responsibility for chain-related effects. Three indicators address the sustainability issue: (i) land use; (ii) water use; and (iii) energy use. When data become available, future studies can establish the interaction between these indicators leading to optimization within environmental constraints varying among regions and in time rather than reduction strategies addressing only one sustainability characteristic. Future research should develop supplemental indicators derived from the core indicators proposed here. This additional set can be used for relevant business sectors. Although the method to assess environmental sustainability was developed for food production, the concept has a much wider applicability.

21

CChhaapptteerr 33

AA mmeetthhoodd ttoo ddeetteerrmmiinnee llaanndd rreeqquuiirreemmeennttss rreellaattiinngg ttoo ffoooodd

ccoonnssuummppttiioonn ppaatttteerrnnss∗∗

Abstract Food production requires agricultural land. The area needed to feed a population depends on the one

hand on production systems (resulting in a specific yield per hectare), and, on the other hand, on the size and the

consumption pattern of this population. Due to various reasons, the amount of land available for food production is

declining. Over the last decade, several studies have published on the issue of food security. Food consumption

patterns, however, also have effects on total land requirements, probably even in the same order of magnitude

than increasing production levels or world population growth. Available studies have estimated that an affluent diet

requires more than three times as much land as a vegetarian diet. This chapter assesses the impact of

consumption patterns on land requirements for food. It develops a method to calculate land required to produce

individual foods. In combination with data on consumption, it determines household land requirements for food.

Applied for the Dutch situation in 1990, the method presents land requirements for over a hundred foods. The

chapter observes large differences among specific land requirements for individual foods. It shows that especially

consumption of livestock foods, oils, fats, and beverages have large effects on household land requirements. Data

on land requirements for specific foods obtained can be used to study the impact of changes in food consumption

patterns. It is stressed that results are typical for the Netherlands in 1990, and can therefore not be used to derive

land requirements for other populations. However, the method can be applied for other countries for which

required data are available. In this way, the thesis contributes to the discussion on future land use. Available

studies on food security have shown that future populations can be fed. The method presented here is a tool for

the evaluation of options to do so in a way that not only physical but also cultural and emotional requirements are

met.

33..11 ►► IInnttrroodduuccttiioonn Up to now, agricultural studies on land use have focused on food security (e.g. Penning de Vries et al., 1995; Bouma et al., 1998a; Groot et al., 1998). History has shown that technological developments in combination with increased global food trade have been the reason that agriculture has been capable to keep up with population growth (Ivens et al., 1992). According to the medium-fertility scenario of the United Nations, the world population will grow from 5.7 billion persons in 1995 to 9.4 billion in 2050 (United Nations, 1998). When high external input farming (HEI) is practiced, simulation of agricultural potentials for the year 2040 indicate that agriculture can feed future world populations (Penning de Vries et al., 1995; Bouma et al., 1998a; Groot et al., 1998). However, due to ongoing industrialization, urbanization, infrastructural developments, land degradation, and desertification, agricultural land resources become more and more scarce. Globally, total, potential agricultural land decreases at a rate of seven percent per decade (Oldeman et al., 1991). Although the per capita world grain harvest increased from 247 to 342 kg (Brown et al., 1997) in the period 1950-1984, the per ∗ This chapter is a slightly adapted version of Gerbens-Leenes, P.W., Nonhebel, S., Ivens, W.P.M.F., 2002. A method to determine land requirements relating to food consumption patterns. Agriculture, Ecosystems & Environment, 90, 47-58.

Chapter 3 __________________________________________________________________________________________

22

capita available agricultural area dropped from 1.5 to 0.8 ha in the period 1961-1998 (FAO, 2001). Moreover, increasing food demand on the world market and the impact of environmental awareness of consumers puts an additional strain on the ability of agriculture to produce enough food in the next decades (Bouma et al., 1998b). On the one hand, production systems, e.g. yields per hectare and efficiency in the food industry, determine land requirements for food; on the other hand, population size and food consumption patterns. Food consumption patterns are repeated arrangements in the consumption of food by a population group and have to do with types and quantities of foods and their combination into different dishes and meals. Tradition and religious rules are often important (Ivens et al., 1992; Whitney and Rolfes, 1999). Food consumption patterns depend on several factors, such as personal preference, habit, availability, economy, convenience, ethnic heritage, tradition, culture, and nutritional requirements (De Wijn, 1971; Von Braun, 1988; Wandel, 1988; Musaiger, 1989; Von Braun and Paulino, 1990; Vringer and Blok, 1995; Van der Boom-Binkhorst et al., 1997; Whitney and Rolfes, 1999). Consumption patterns, therefore, differ among communities and generations (Jobse-vanPutten, 1995). To provide enough nutritional energy to stay alive, food has to supply about 10 MJ (2400 kilocalories) per capita per day (Voedingscentrum, 1998a). However, as soon as welfare rises above the subsistence level, people put other demands on food than ‘enough to stay alive’. Features like taste, day-to-day variety, and convenience also become important. Dutch environmental studies have shown that in the past decades, increasing incomes resulted in a larger use of natural resources such as water (Achtienribbe, 1993) and energy (Vringer and Blok, 1995). Despite the physiological limits of food intake, there could well be a relationship between increasing affluence in society, and hence more affluent food consumption patterns, and larger claims on agricultural land. Available studies have estimated that an affluent diet requires more than three times as much land as a vegetarian diet (Penning de Vries et al., 1995; Bouma et al., 1998a; Groot et al., 1998). This implies that the impact of food consumption patterns on land required is probably in the same order of magnitude than changing production levels and the growing world population, but up to now this impact has not been studied in a quantitative manner. This chapter studied the relationship between food consumption patterns and land requirements. The chapter presents a method for assessing land requirements for food that requires detailed microlevel information on land requirements for individual food items and household consumption. Large inter-generational and regional differences among food consumption patterns, as well as yield differences among regions, however, complicate the analysis. For example, large differences in the consumption of specific foods occur among the countries of the European Union (LEI-DLO/CBS, 1998; FAO, 2001). Another complication is that factors that determine land requirements are interrelated. Systems provide the availability of commodities, but demand determines what and how much is produced, and the waste streams that are generated. This implies that the calculation of land requirements can only be performed for a clearly defined situation. This chapter applied the method to assess requirements for Dutch households. Firstly, it calculated land requirements (m2 year kg-1) for over a hundred commonly used food items in the Netherlands. Secondly, in order to assess household land requirements for food, it combined these data with data on household consumption (kg year-1). Information obtained at the bottom of the food system, the household level, represents a major improvement in the quality of the analyses on land required for food. This type of information provides a valuable link to how food consumption patterns aggregate demand of resources. The method provides a tool to study the impact of changes in consumption patterns on land requirements. In this way, it contributes to the discussion on future land requirements. 33..22 ►► SSyysstteemm ddeessccrriippttiioonn Agricultural production provides the availability of food commodities. The food industry processes these commodities in various ways for the production of an enormous variety of food and non-food items. Food items or foodstuffs are products of animal or vegetable origin that are available for consumers. They form the end of a food production chain. The production of foods in these chains requires several basic ingredients, originating from different food production systems, and with different land requirements. The production of cakes, for example, requires a number of basic ingredients, such as sugar, flour, eggs, and butter, ingredients that all have different land requirements. Land requirements for individual food items, therefore, can differ considerably. Per capita available amounts of food, like grain or soyabeans, are well known (Brown et al., 1997). However, what production systems do with these commodities is less clear. Consumers do not eat grain, they eat bread, pasta, or meat. The food industry can directly process grain or soyabeans to

A method to determine land requirements relating to food consumption patterns

__________________________________________________________________________________________

23

produce food items, like bread or oil. It can also apply grain or soyabeans for livestock fodders, so that basic ingredients are used indirectly to produce livestock food items, like meat, milk, or eggs. Consumption patterns, i.e. demand for food on a household scale level, therefore, determines what production systems do with available commodities. In this way, demand for specific foods strongly influences the size of the required agricultural area. Food production systems show large complexity. This thesis made a rough division between primary and secondary production systems. Primary production grows crops that form the basis for secondary or livestock production. Land requirements for commodities originating in primary production systems depend on yields, and therefore on crop characteristics and countries of origin. Yields depend heavily on the applied production system. Primary production ranges from systems with high external input and high yields (HEI), and systems with low external input and relatively low yields (LEI). Livestock production systems use large amounts of wastes originating in the food industry. The production of soya oil, for example, generates large amounts of oil cakes that are used for livestock fodders. Land requirements for basic ingredients originating in secondary production, therefore, depend on: (1) supply of plant materials and waste streams; (2) animal species; and (3) production system characteristics. On a national level, countries provide in their food need either by importing commodities or produce them in the country itself, many countries export commodities. 33..33 ►► MMaatteerriiaallss aanndd mmeetthhooddss Although several approaches can assess total land requirements for food, every method must combine data on food consumption with data on production. Environmental research has developed a method to calculate energy requirements for household purchases (Van Engelenburg et al., 1994) and for food items (Kok et al., 1993). That method uses items of consumption as its basic unit of measurement. The assessment of per household consumption is based on household expenditure in combination with prices of consumption items. Those studies combine results on household consumption with calculations on energy requirements for consumption items. This thesis adopted the method for the assessment of land requirements for food. What is finally arrived at by combining household consumption (kg) with land requirements per food item (m2 year kg-1) is the land requirement for food at a household level. The thesis considered over a hundred food items divided into nine main consumption categories: 1. meat; 2. dairy and eggs; 3. beverages; 4. cakes and pastries; 5. potatoes, vegetables and fruits; 6. oils and fats; 7. bread; 8. flour products, and 9. ‘other food products’. To determine land requirements for food items, the thesis needed information on yields, imports, food industry recipes, and proportions of crops grown in the open air and in glasshouses. For the Netherlands, information on yields is available on different levels of scale (FAO, CBS, IKC-AG). Information on imports is available from the Dutch Bureau of Statistics (CBS). The thesis derived information on recipes in the food industry and data on the proportion of crops grown in the open air and in glasshouses from knowledge obtained in energy studies (Kramer and Moll, 1995). 3.3.1 Scale levels for land requirements for food For the assessment of land requirements for food, the thesis distinguished five scale levels. Figure 3.1 shows these levels: (1) the primary production level; (2) the secondary production level; (3) the food item level; (4) the household level, and (5) the national level. Data from a lower level formed the input data for a higher level. The thesis calculated the average land requirement for the household (level 4) by combining data from level 3 with data on household consumption. This land requirement can be scaled up to the national level by multiplying the household land requirement by the total number of Dutch households (level 5).

Chapter 3 __________________________________________________________________________________________

24

3.3.2 Flow chart of calculations and input The thesis determined the total land requirement for food in the Netherlands in a step-by-step approach in which the output of one step formed the input of the next one. Figure 3.2 shows the flow chart, the seven steps, and necessary inputs.

Fig. 3.1. Schematic representation of the 5 levels of scale for the land requirements for food. Level 1: the primary production level; level 2: the secondary producton level; level 3: the food item level; level 4: the household level and level 5: the national level. For the assessment of level 4, data from level 3 are coupled with data on household consumption (lr = land requirement).

Step 1 assessed prices of food items in 1990 using national statistical data (CBS, 1980, 1986, 1991), or data derived from a study on energy requirements for food (Kramer and Moll, 1995). Step 2 determined the amount of food items bought by an average Dutch household by combining prices of food items along with household expenditure. The thesis obtained data from the Dutch Household Expenditure Survey (CBS, 1990) which is based on a representative sample of households. Step 3 made a calculation of land requirements for Dutch crops. Land requirements for crops are inversely proportional to the yield. It is expressed as area per kg crop year (m2 year kg-1). For each agricultural crop, the thesis calculated land requirements by dividing the total Dutch cultivated area (m2) by the total Dutch yield per year (kg year-1). The thesis obtained data from CBS-inventories (CBS, 1990, 1993). Vegetables and fruits are cultivated either in the open air or in glasshouses. For open air vegetables and fruits, and for tomatoes, capsicums, and cucumbers in glasshouses, the thesis calculated land requirements by dividing the total cultivated area per crop by the Dutch yields of 1990. For the latter assessment, it derived data from CBS-inventories (CBS, 1993) and from inventories of the ‘Informatie- en Kenniscentrum voor de Akkerbouw en de Groenteteelt in de Vollegrond’ (IKC-AG, 1993). Vegetables and fruits grown in glasshouses have several yields per year, however. The thesis assessed total yields (kg m-2 year-1) by multiplying yields per production round (Ekkes et al., 1994) by

lr. food item lr.food item lr. food item lr. food item

lr. basic material

lr. basic material

lr. basic material

lr. basic material

lr. basic material

lr. basic material

level 4: household level

level 3:food item level

level 2: secondary production level

level 1: primary production level

level 5: national level

national lr.

lr. livestock products

household lr.


__________________________________________________________________________________________

25

the estimated number of production rounds per year. From these annual yields, it calculated land requirements. Finally, it calculated weighed averages of land requirements for Dutch vegetables and fruits grown in the open air and in greenhouses. Step 4 calculated weighed average land requirements for Dutch and imported crops. The thesis calculated land requirements for crops available in the Netherlands in two phases. Step 4a assessed the weighed average land requirements of imported crops. The estimated amount and origin of foreign crops derived from inventories of the LEI-DLO and the CBS (LEI-DLO / CBS, 1991). The annual yields of imported crops derived from FAO-inventories (FAO, 1990, 2001). Step 4b calculated the weighed average land requirements for crops available in the Netherlands. Using land requirements for Dutch crops from step 3, this was done by assessing the weighed average of the Dutch and the imported crops. Fig. 3.2. Schematic representation of steps and inputs (1990) for the assessment of the total land requirement for food in the Netherlands in which the output of one step forms the input of the next one. (a) the base-requirement = food items bought by an average Dutch household in 1990. (b). the base-land requirement = the area needed to produce the amount of food for the base-requirement. Step 5 determined land requirements for livestock food items (it excluded fish). The Dutch livestock production system can roughly be divided into intensive animal husbandry and dairy farming. Livestock is fed with roughage and/or feed-concentrates. In intensive animal husbandry, the production of livestock fodder does not take place at the farm itself, while in dairy farming domestic animals are mostly fed with roughage obtained from the farmers own agricultural production system. Roughage is fresh or dried fibrous fodder, such as grass, silage, fodderbeets, potatoes, and straw. Feed-concentrates are made of crops (mainly grains, pulses, lupines, and tapioca), and by-products and wastes originating in the food industry. The composition of feed-concentrates depends heavily on prices of commodities on the world market, and is therefore constantly changing (Kingmans, 1998).

Step 3 land requirements

Dutch crops

land requirements imported crops,

imports, proportions open air and

greenhouse crops

Step 4 weighed average land

requirements

Step 6 land requirements food

item

Step 1 prices food items

Step 2 base requirement

(a)

Step 7 base-land

requirement (b)

Step 8 total land requirement for food

in the Netherlands

cultivated area, total Dutch harvests,

number of harvests

Dutch meat and milk production,

composition livestock fodders, data

livestock

expenditure for food

prices ’80,’85 price index ’85,’90

Step 5 land requirements livestock products

recipes food industry

INPUTS STEPS INPUTS

persons per household, number of

inhabitants

Chapter 3 __________________________________________________________________________________________

26

The thesis calculated land requirements for livestock food items by dividing the areas needed for fodders (m2 year) by the yearly production of meat, raw milk, and eggs (kg). The thesis assigned land to the main product, no land requirements were assigned to by-products and wastes from the food industry that are used in the fodders of domestic animals. It derived data on livestock, livestock fodders, and livestock products from the ‘Yearly statistics of livestock fodders of 1990/’91 and 1991/’92’ (Bolhuis et al., 1995), from announcements of the ‘Marketing Board for Livestock, Meat and Eggs’ and from CBS-inventories (CBS, 1993). Some livestock production systems make only one food commodity, while other systems produce two. Pig keeping, for example, results in the production of only pork, whereas dairy farming produces raw milk as well as beef. When a production system makes more products, the thesis divided the total land requirement over the products proportional to their energy output (kJ). Raw milk is the basic ingredient for the production of various food items, like semi-skimmed milk, yoghurt, cheese, and butter. In order to calculate land requirements for these foods, the thesis assessed land requirements for the components milk carbohydrate, -fat and -protein. This was done by dividing the land requirement for raw milk (m2 year kg-1) over carbohydrates, fats, and proteins according to their energy content (kJ kg-1). Next, the thesis assessed land requirements for dairy products according to their composition. Step 6 assessed land requirements for specific food items. The food industry applies basic ingredients to produce food items. The thesis calculated land requirements for food items by multiplying land requirements for these basic ingredients (m2 year kg-1) by the amounts needed to produce the food item (kg kg-1) and summing these results. It derived the amounts of basic ingredients needed for the manufacturing of food items from Kramer and Moll (1995), and data on ingredients used in the food oil industry from inventories of the ‘Marketing Board for Margarines, Fats and Oils’ (Produktschap Margarine, Vetten en Oliën, 1993). It took data on the composition of food items from the ‘Dutch food items table’ (Voorlichtingsbureau voor de Voeding, 1984, 1990). Step 7 determined the land requirement for food for an average Dutch household by combining data on the amount of food items bought by an average household in 1990 from step 2 with land requirements for specific foods per food category from step 6. Step 8 assessed the total land requirement for food in the Netherlands in 1990 by multiplying the base-land requirement calculated in step 7 by the total number of households in the Netherlands (CBS, 1993). 33..44 ►► RReessuullttss aanndd ddiissccuussssiioonn 3.4.1 Land requirements The thesis obtained results on several scale levels. Step 2 provided average amounts of food items bought by a Dutch household in 1990, step 6 land requirements per food item. Table 3.1 shows some of these results for commonly used foods. Appendix A presents an overview of all results. For individual food items, large differences existed among household consumption and among land requirements. In general, more expensive food items had larger land requirements. For example, in 1990 the price of beef was f 17.00 kg-1, of pork only f 11.00 kg-1 (Kramer and Moll, 1995). The land requirement for beef was 20.9 m2 kg-1 year, more than twice the land requirement for pork (8.9 m2 kg-1 year). This was partly due to a larger conversion efficiency of swine (energy in produce/energy in feed = 0.35), while the conversion efficiency of beef is only 0.06 (Spedding, 1988). The second reason was that 47% of the energy content of pigfodders was provided by wastes and by-products originating in the food industry, while for beef this amount was only 27% of the total energy content (Bolhuis et al., 1995). As mentioned before, the thesis assigned no land requirements to these wastes and by-products. The household land requirement for a specific type of food item is determined by the amounts bought by an average household, as well as by the individual land requirement for that type of food item. Some food items showed large land requirements but low consumption, resulting in relatively low land requirements per household (e.g. tea). Other food items showed small specific land requirements but relatively large household land requirements due to large consumption (e.g. semi-skimmed milk and potatoes). In the Netherlands in 1990, due to relatively large individual land requirements in combination with consumption, only six food items accounted for 42% of total household land requirements. These items were: (1) margarine (13%); (2) cheese (8%); (3) minced meat (7%); (4) sausages (5%); (5) fats for frying (5%); and (6) coffee (4%). Staple foods, like potatoes, vegetables, fruits, bread, and flour products accounted for only 12% of the total Dutch household land requirement.


__________________________________________________________________________________________

27

Table 3.2 shows land requirements per consumption category for an average household. The category of meat was the largest category with 1022 m2 (29% of the total). Of the total land requirement for an average household, oils and fats needed 827 m2. This land requirement was based on the use of soya oil in the food oil industry. The assessment of land requirements for individual food items was often complicated by a joint production dilemma, however. For soyabeans, all of the land required was classified as being for soya oil, while oil cakes used for livestock fodders were merely seen as being a waste product. The consequence of assigning all the land to the main product was that land requirements for oil products were relatively large, while land requirements for livestock foods were relatively small. Table 3.1 Specific land requirements per food item (m2 year kg-1) based on local yields and the Dutch production situation in 1990 and household food consumption in the Netherlands (kg year-1) based on the CBS Household Expenditure Survey of 1990

Food item Specific land requirement (m2 year kg-1)

Household consumption(kg per year)

Beverages Beer 0.5 77.3 Wine 1.5 29.3 Coffee 15.8 8.9 Tea 35.2 1.8 Fats Fats for frying 21.5 8.0 Margarine 21.5 20.6 Low fat spread 10.3 9.8 Meat Beef 20.9 8.4 Pork 8.9 15.1 Minced meat 16.0 16.0 Sausages 12.1 14.9 Dairy and eggs Whole milk 1.2 34.6 Semi-skimmed milk 0.9 116.1 Cheese 10.2 27.6 Eggs 3.5 13.4 Cereals, sugar, potatoes, vegetables and fruits

Flour 1.6 9.4 Sugar 1.2 22.3 Potatoes 0.2 153.6 Vegetables (average) 0.3 162.0 Fruits (average) 0.5 154.0

The land requirement for the consumption category of dairy and eggs was 598 m2 for an average household; 17% of the total land requirement for food in the Netherlands. Raw milk is the basic ingredient for various dairy products. Because of the chosen method, the land requirement for milk fat (16.6 m2 kg-1) was twice as large as requirements for milk-protein and -carbohydrates (7.4 m2 kg-1). As a consequence, land requirements for dairy products were dominated by their fat content. Compared to meat, the land requirement for eggs was relatively small (3.5 m2 year kg-1). Of the total land requirement for a household, 368 m2 (11%) was needed for beverages. More than half of this requirement, 203 m2, was needed for coffee and tea. The area needed for these two beverages was even larger than land requirements for basic food items like bread, potatoes, vegetables, and fruits. In reality, however, the amounts of coffee consumed were even larger than the amounts calculated in

Chapter 3 __________________________________________________________________________________________

28

this thesis. According to the ‘Association of Dutch Coffee Roasters and Tea Packers’, the total amount of coffee consumed per capita in 1990 was 8.4 kg (Vereniging van Nederlandse Koffiebranders en Theepakkers, 1998). This thesis calculated only 3.7 kg per capita. The difference between average household and actual consumption indicated that most of the coffee is not consumed at home, but at work or in restaurants. Table 3.2 shows that only a relatively small area is needed for the categories of bread; potatoes, vegetables and fruits; cakes and pastries; and flour products. Results showed that even small changes in food consumption patterns can have large impacts on the agricultural area required to produce this food. For example, in the Netherlands a hot meal mostly includes some meat; potatoes, rice or pasta; and vegetables. A slight increase of the consumption of meat by only one mouthful (10 grams) per capita per day will increase the agricultural area required by 103 m2 per household per year (+ 3%), whereas the same increase of potato consumption will result in an increase of only 2 m2 per household per year (+ 0.05%). Table 3.2 Land requirements (m2 per household and % of the total agricultural area for household food consumption) for the 9 main consumption categories (meat; oils and fats; milk products and eggs; beverages; bread; potatoes, vegetables and fruits; cakes and pastries; flour products and other food products) based on the Dutch production situation and household consumption in 1990, and the total land requirement per household for food consumption

Consumption category Land requirement (m2 per householda)

Land requirement (% of the total agricultural area for household food

consumption) Meat 1022 29 Oils and fats 827 24 Dairy and eggs 598 17 Beverages 368 11 Bread 183 5 Potatoes, vegetables and fruits 168 5 Cakes and pastries 107 3 Flour products 68 2 Other food products 148 4 Total 3490 100

a. In 1990 an average Dutch household consisted of 2.41 persons (CBS, 1993).

3.4.2 Effect of uncertainty and inaccuracy on final results For the assessment of land requirements for food, the thesis required information from several sources. On the one hand, all information obtained shows uncertainties that are related to the impossibility to do an accurate description of a system. For example, uncertainties related to food consumption are caused by losses in the life cycle, outdoor consumption, and non-food purposes. In the life cycle of a food item, about 10% of the total amount is lost (Groenendaal, 1996). The CBS expenditure survey does not take consumption outside the house into account. For example, as mentioned before, the thesis found large differences between average, household coffee consumption and the availability of coffee on a national scale. Most agricultural crops are used for food purposes, but some are applied in other areas. Potato starch, for example, is used for foodstuffs and animal feed, but most of the starch is used for non-food purposes (LEI/CBS, 1992). On the other hand, consumption data are inaccurate because of underreporting and price effects. The CBS has concluded that households underreport their expenditure (CBS, 1994). The CBS has estimated that underreporting is about 16% of total, household expenditure. Some expenditures are presumed to be even more underreported. In an earlier study on energy requirements for food items, it was argued that underreporting of alcoholic beverages is about 34% of the total (Kok et al., 1993). The CBS assesses prices of food items by using a fixed package of products. In this assessment, weekly offers of supermarkets are not taken into account. This means that the use of these food price data underestimates real amounts of food items bought by households. In this thesis, it is argued that the amounts of food bought by an average Dutch household were 10% higher than the amounts of the base-requirement because of this effect. In order to calculate the sensitivity of the model, the thesis took underreporting, price effects, losses in the life cycle, and outdoor consumption cumulatively into account. The land requirement for food for


__________________________________________________________________________________________

29

an average, Dutch household lied between 3490 and 5243 m2. Uncertainty and inaccuracy in available consumption data all increased household land requirement, so that actual requirements were probably close to the maximum of the calculated interval. This is why the thesis did not give average value. For some foods, underreporting is probably larger than for others (e.g. alcoholic beverages), while other food items are mostly consumed outside the house (e.g. coffee). These inaccuracies and uncertainties influenced the fractions of household land requirements per consumption category, so that some fractions were probably larger than calculated (e.g. beverages), while others were smaller (e.g. bread). 3.4.3 Comparison with data on available food on a national level To be able to compare results with calculations on a national scale, land requirements at a household level had to be scaled up to the national level. Based on household consumption, the total Dutch land requirement for food lied between 2.2 x 106 and 3.2 x 106 ha. In the Netherlands, data on consumption are available from inventories on a national scale. Every year, the Dutch Agricultural Economic Institute (LEI) and the Central Bureau for Statistics (CBS) publish data on the amount of food available in the Netherlands (Landbouwcijfers). In order to study the impact of assumptions on the final result, the land requirement for the available food in the Netherlands was also calculated based on these LEI/CBS inventories. This requirement was 3.0 x 106 ha, which lies in the interval calculated above. 3.4.4 Application and sensitivity of the method Results showed that household consumption strongly determined the total land requirement for a populations food need (Tables 3.1 and 3.2). Especially the consumption of livestock foods (dairy and meat), vegetable oils and fats, and beverages had a large effect on land requirements for food. The evaluation of results obtained in this thesis showed that Dutch land requirements for meat and milk accounted for nearly half of the total land requirement for food, while land requirements for bread, potatoes, vegetables, fruits, and flour products accounted for only 12% of the total. A shift to a more luxury diet often implies the consumption of more or more luxury meat types. Expensive meat types (e.g. beef) often have larger land requirements than cheaper ones (pork). But not only did changes take place in the category of meat, shifts occur in all food categories: other vegetables, more fruits, more snacks and cakes etc. All these more luxury food items require more land. For example, Dutch land requirements for beverages were large: 11% of the total land requirement for food was needed for the production of beverages, mostly for coffee. The so-called non-meat changes in the menu (oils, beverages, fruits, cakes etc.) seem to have a large impact on land requirements. The effect of these types of consumption changes on land requirements can be studied via the route presented in this chapter. It is stressed, however, that the method provides a tool to calculate slight demand changes using the land requirements for individual food items generated, but that larger changes (e.g. a shift towards a vegetarian menu) make new assessments necessary because these changes will influence the whole food production system. For example, if the consumption of fats rises, this generates larger waste streams. These streams can be used for livestock production, so that land requirements for meat, eggs, and milk will drop. Penning de Vries et al. (1995) and Bouma et al. (1998a) have estimated that a shift from a vegetarian diet to an affluent diet with meat leads to a tripling of the land requirement. Based on an average energy requirement of 10 MJ per capita per day, the consumption of wheat, and data obtained in this thesis, the hypothetical land requirement for a household at the subsistence level is 444 m2, a factor of eight smaller than the average Dutch household land requirement calculated in this thesis. This gives an indication that the difference between a vegetarian and an affluent diet with meat could be much larger than the factor of three estimated by the authors mentioned above. The method presented here can only be used to assess land requirements for a clearly defined food production and consumption system. Land requirements for food items are strongly related to yields per hectare. Halving the yields leads to doubling the land requirements. The production situation in the Netherlands, for example, is characterized by high yields per hectare in comparison with average, global yields, but also with yields in other European countries. In 1996, the average, Dutch wheat yield was 8.9 tons per hectare, while the global average was only 2.5 and wheat yields in Italy 3.3 tons per hectare. This implies that if a production based on average, global yields has to meet Dutch consumption, this will require about three to four times as much land. These yield differences are the main reason for large variations that occur in various footprint studies. Wackernagel et al. (1997) have assessed the footprint for Dutch food consumption. They have used global yields as input data and

Chapter 3 __________________________________________________________________________________________

30

have arrived at 33 x 106 ha (including 7 x 106 ha for fish), one order of magnitude larger than results of the current thesis. Imports and exports of considerable quantities of commodities and food items on a world scale influence statistical data of individual countries. If assessments of land requirements for food take only imports into account, they will overestimate requirements for countries with highly productive food industries, and consequently large exports, like for example the Netherlands. Land requirements can be assessed using food balance sheets, determining land use due to imports, domestic land use, and subtracting land use for exports. This method has been used by the RIVM (Van Vuuren et al., 1999). A disadvantage is that the method does not take processes in the food industry into account. Imported grain, for example, can be used for the production of bread, but also for the production of pork for export purposes. The current thesis assessed land requirements for individual food items, so that in combination with data on consumption, it calculated only land requirements in relation to consumption patterns of the population considered, and avoided double countings. Two Dutch studies on food security in the Netherlands (Bakker, 1984; Groenendaal, 1996) have assessed land requirements based on Dutch yields and a minimum consumption. Those studies have shown that food consumption in the Netherlands could be provided for with present yields on the available area for agriculture (2 x 106 ha) which is in the same order of magnitude than results of this thesis. However, the studies were less detailed than the present one, so that they could not give the effect of slight changes of the system. The sensitivity of results of the method to yields per hectare, in combination with properties of the food system, implies that values for food items derived in this thesis cannot be used to assess land requirements related to consumption patterns in other situations. Yield levels, production systems, and consumption patterns change in time and vary among populations, so that data obtained here are only valid for the Netherlands in 1990. In most other countries, yield levels are lower, so that land requirements for food items will be larger. It is stressed that results obtained by the method can not be used as an indicator for sustainable land use. Comparing land requirements for different systems is complicated by the fact that productivity is not only influenced by human management factors, but also by natural factors, such as climate and soil quality. Agricultural systems range from systems with high external input and high yields to systems with low external input and relatively low yields. The present Dutch agricultural system is a so-called high input system focused on high yields per hectare, requiring large input of chemical fertilizers and biocides leading to large emissions to the environment. Increased environmental awareness might lead to the introduction of more environmentally friendly agricultural systems, often achieving lower yields per hectare. This may imply larger land requirements in the future. Data on land requirements for individual food items, in combination with microlevel information on household consumption obtained at the bottom of the food system, can give a valuable contribution to the discussion on land requirements for populations. The method presented here provides a tool to study other combinations of consumption patterns and production systems, for example, other countries, or future perspectives. This requires other information on agricultural yields, consumption patterns, food production systems, etc. This type of information is available for nearly all developed countries. 33..55 ►► CCoonncclluussiioonnss Results of the current thesis showed that an analysis of the land requirement for food based on data on food production and on household consumption obtained at the top and at the bottom of the food system provides new perspectives. When the method was used to assess requirements for Dutch households, the thesis showed large differences among land requirements for the nine consumption categories defined, and even among individual food items of the same category. Almost half of the Dutch household land requirement for food is needed for only six food items: margarine, cheese, minced meat, sausages, fats for frying, and coffee. Staple foods, like potatoes, bread and flour products, and vegetables and fruits, accounted for only 12% of the total household land requirement. Changes in food consumption patterns will have large effects on total land requirements for food and are in the same order of magnitude than changing production levels or the growth of the world population. The current thesis described only 72% of Dutch food consumption, but microlevel information provided by this analysis can form a basis for further research into ways of changing land requirements by adding the perspective of the household and food item level. Data obtained with the method developed are only valid for the Netherlands in 1990, but provided the necessary information is available, other countries can adopt the method. In this way, the thesis presented here can


__________________________________________________________________________________________

31

contribute to the discussion on the ability to produce enough food in the next decades. Available studies on food security have shown that future generations can be fed. The method presented in this chapter provides a tool to assess the possibility to do so in a way that not only physical but also cultural and emotional requirements are met.

32

33

CChhaapptteerr 44

CCoonnssuummppttiioonn ppaatttteerrnnss aanndd tthheeiirr eeffffeeccttss oonn llaanndd rreeqquuiirreedd ffoorr ffoooodd∗∗

Abstract Food production requires vast amounts of land but the area suitable for growing crops is limited,

however. This chapter pays attention to the relationship between food consumption patterns and agricultural land

requirements. Land requirements per food item that were determined in chapter two are combined with data on

per capita food consumption of various food packages, varying from subsistence to affluent, leading to information

on land requirements for food. The thesis showed that there are large differences among per capita food

consumption patterns and related land requirements, while food consumption, expenditure, and the physical

consumption of specific foods change rapidly over time. It found a difference of a factor of two between

requirements for existing European food patterns, while the land requirement for a hypothetical diet based on

wheat was six times less than that for an existing affluent diet with meat. It is argued that in the near future

changes in consumption patterns rather than population growth will form the most important variable for total land

requirements for food. Trends towards the consumption of foods associated with affluent lifestyles will bring with

them a need for more land. Lifestyle changes, changes in consumer behavior on a household level, can be

considered as powerful options to reduce the use of natural resources such as agricultural land.

44..11 ►► IInnttrroodduuccttiioonn Suitable soil is a limited, necessary resource for the production of food. On a global scale, 31% of the soil surface can be used for arable crops, while an additional 33% is suitable for grassland (Penning de Vries et al., 1995). High quality arable land is becoming scarcer and scarcer due to ongoing industrialization, urbanization, infrastructural development, land degradation, and desertification (Oldeman et al., 1999). Land requirements for food are, among other things, determined by population size, and by the types and amounts of specific foods consumed, i.e. food consumption patterns. The discussion of world agricultural futures has usually been framed in a Malthusian context, with technological optimists opposing neo-Malthusian pessimists (Harris, 1996). The Second World War marked a turning point in the yield per hectare of arable crops in the Western world. Before World War ll, for example, wheat yields in the United Kingdom and the USA increased only by a few kg ha-1 year-1 (De Wit, 1992). As a result of the first “green revolution”, yields have consistently increased at much higher rates. The continued increase in production per unit of land area as well as per unit of livestock has led to significant increases in agricultural productivity (Rabbinge and Van Latesteijn, 1992). Over the last decades, several studies on agricultural potentials have come to the conclusion that modern agriculture can theoretically provide enough food to feed the world’s growing population (Penning de Vries et al., 1995). Existing studies on food security indicate that agriculture can feed future generations, while surpluses of certain commodities in the European Union give the general impression that more than enough agricultural land is available. The surplus could then be used for other purposes, such as cultivation of energy crops, development of new nature areas, ecological forms of agriculture, or infrastructural developments. A basic food function is to provide enough energy and nutrients for body functions and physical activity, resulting in a large variety of menus that meet nutritional constraints. Calculations about food security have focused on these basic functions. However, once people can fulfill their physiological

∗ This chapter is a slightly adapted version of Gerbens-Leenes, P.W. and Nonhebel, S., 2002. Consumption patterns and their effects on land required for food. Ecological Economics, 42, 185-199.

Chapter 4 __________________________________________________________________________________________

34

requirements, and, at the same time, economic resources are available, social and cultural aspects of food become important, resulting in actual food consumption patterns that are much more varied than menus on the basic level. Increasing production can guarantee food security, but it does not guarantee a sufficient availability of all the foods needed to satisfy consumer demand because of additional functions of food. With respect to food, a distinction can be made between physical consumption and consumption in the form of expenditure. It should be realized that there is a physiological limit to consumption. To provide energy for body functions, requirements are still in the same order of magnitude that they were in the Stone Age, which is about 10 MJ (2400 kilocalories) per capita per day (Voedingscentrum, 1998a). Economic consumption, however, can rise almost infinitely. Many studies have shown that the overall composition of people’s diets corresponds to their income (Von Braun, 1988; Vringer and Blok, 1995). In general, when standards of living are low, increasing incomes will favor more foods of animal origin, while the consumption of grains and carrots will drop (Grigg, 1994). Beyond basic constraints, rising incomes do not favor more food but rather more expensive foods, so that dietary shifts and not increasing physical consumption are responsible for changing claims on available natural resources. In the Netherlands, for example, increasing affluence in society has resulted in a substantial rise in meat consumption over the last decades, with per capita consumption rising from 36 kg in 1950 to 90 kg in 1990 (LEI, 1978, 1993). Agricultural studies on food security (Penning de Vries et al., 1995; Bouma et al., 1998) have estimated that a shift from a vegetarian diet to an affluent diet with meat leads to a threefold increase in the land required. Results of a study on land requirements for food (Gerbens-Leenes et al., 2002) have indicated that the difference between an affluent diet and a vegetarian one could even be larger than the factor of three estimated by the authors mentioned above. However, affluent diets, such as the present diets in the countries of the European Union, not only imply the consumption of more or other types of meat, shifts occur in all food categories. The so-called non-meat changes in the menu (more oils, beverages, fruits, cheese, ice cream, cakes, and so on) seem to have a large impact on land requirements because, in general, more affluent foods require more land. For example, the land requirement (m2 kg-1) for beef, which is relatively expensive, is more than twice the requirement for pork (Gerbens-Leenes et al., 2002). The studies mentioned above indicate that dietary change rather than increasing physical consumption is responsible for larger claims on the available agricultural area. Nowadays, per capita food consumption shows large differences between developed and developing countries. In the world's poorest countries, average food intakes are even too low to prevent malnutrition and hunger (Azoulay, 1998). Studies on food security indicate that agricultural production is capable of securing physiological requirements, but they also show that an affluent diet with meat can probably not be produced for the total world population (Penning de Vries et al., 1995). In the near future, the social and cultural aspects of food will therefore be of crucial importance for satisfying food demand on a global scale. The specific aims of this chapter are to examine quantitatively the range of per capita land requirements for food related to food packages that not only fulfill basic physiological needs, but also satisfy social and cultural demands. The chapter calculates inter-generational and regional differences in consumption, and identifies those parts of the food packages that have the largest claims on the required land area. The results are used to discuss future food security on a global scale. The paper is organized as follows. First, it describes food consumption patterns and the factors that determine them. A distinction is made between menus fulfilling the basic functions of food and menus that also meet the needs of social and cultural demands. Next, the paper presents a method that combines available data on food consumption with data on specific land requirements for foods (Gerbens-Leenes et al., 2002). On the basis of this, it assesses the gap between land requirements for basic physiological food needs on the one hand, and land requirements for existing food consumption patterns on the other. So far, when calculating land requirements, studies on food security have only considered differences in meat consumption. However, there are more foods requiring substantial amounts of agricultural land. Therefore, this thesis not only calculates the influence of meat on the land required, but also of complete food packages, including the effect of beverages. By assessing the resource costs of various food consumption options, this thesis makes a contribution to the current discussion on future land requirements.

Consumption patterns and their effect on land required for food

_________________________________________________________________________________________

35

44..22 ►► FFoooodd ccoonnssuummppttiioonn ppaatttteerrnnss There is a link between food consumption and nutritional status. However, nutritional status is not a direct reflection of the intake of foods, but rather of the energy and nutrients provided by the foods (Ivens et al., 1992). The edible parts of both animal and vegetable foods consist of water, carbohydrates, lipids, proteins, vitamins, and minerals. The first four are macronutrients; the last two are also referred to as micronutrients, as they occur in small quantities. A distinction can be made between types and amounts of foods needed to provide requirements for nutrients, e.g. vitamins and minerals, and amounts required to provide sufficient nutritional energy. The latter amounts depend on the physical activity engaged in by the person and are in many cases higher than the former quantities (Voedingscentrum, 1998a). To describe the food consumption of a group, one generally tries to discern basic regularities, referred to as food consumption patterns. Food consumption patterns are repeated arrangements that can be observed in the consumption of food by a population group (Ivens et al., 1992). It concerns the types and quantities of foods, and their combinations into different dishes or meals. Food consumption patterns are not static, although it has been found difficult to change them (Ivens et al., 1992). Consumption patterns develop over the course of generations and can differ strongly among communities (Jobse-van Putten, 1995). They depend on several factors, such as personal preference, habit, availability, economy, convenience, ethnic heritage, religion, tradition, and nutritional and cultural requirements (De Wijn and Weits, 1971; Von Braun, 1988; Wandel, 1988; Musaiger, 1989; Von Braun and Paulino, 1990; Ivens et al., 1992; Vringer and Blok, 1995; Whitney and Rolfes, 1999; Van der Boom-Binkhorst et al., 1997). In general, foods are perishable. So, until the introduction of modern transportation and food conservation techniques, such as freezing and cooling, only well-dried food items such as grains, coffee or dried fish could be traded (Jobse-van Putten, 1995). These foods were expensive, so until recently, people were strongly dependent on the local availability of foods. The study of Jobse-van Putten (1995) showed that throughout European history, one of the most typical characteristics of food consumption patterns was the continuous alternation between scarcity and abundance. During the 20th century, consumption patterns have shifted away from traditional food, mainly harvested from the local environment, towards a diet of market food. Gradually, a more varied consumption pattern has developed (Landbouw-Economisch Instituut/Centraal Bureau voor de Statistiek, 1980, 1985, 1996). 44..33 ►► MMaatteerriiaallss aanndd mmeetthhooddss 4.3.1 Starting points This paper distinguished food requirements on three scale levels. These are shown in Figure 4.1. On the first level, the basic level, energy requirements were met by the consumption of wheat. Studies on food security that compare food consumption and potential production express both in grain equivalents (GE) (e.g. Penning de Vries et al., 1995). In the consumption process, GE refer to the amount of cereals needed for the food consumed, plus the 'opportunity cost' to grow food that cannot be produced via grain. This diet will prevent starvation, but will cause malnutrition in the long run because many essential nutrients are lacking. Food requirements on the second level, the subsistence level, are optimal from a nutritional point of view. They are based on a selected number of nutrient-dense foods, providing bodily health for the total life span. Food requirements on the third scale level, the cultural level, form the actual consumption patterns. They also contain foods low in nutrient-density, for example coffee, cakes or chocolate, or larger amounts of foods than requirements on the subsistence level. The total land requirement for a certain type of food is determined by the specific land requirement for that type and by the amounts consumed. This chapter used results of Chapter 3 on land requirements relating to food consumption patterns as the basis for the calculations. That chapter has developed a method to determine land requirements relating to food consumption patterns. It has applied this method for the Dutch situation in 1990, which has resulted in an overview of specific land requirements for over a hundred commodities and food items (m2 year kg-1) available in the Netherlands, including imported foods such as coffee or soybeans that cannot be produced in western Europe. This chapter calculated land requirements (m2 capita-1) for the three scale levels defined above by multiplying consumption (kg capita-1 year-1) per food item by the specific land requirements for that item (m2 year kg-1) and summing the results. Since calculations were based on data of the Dutch food production system 1990, it is stressed that results obtained can not be used to compare the various footprints of nations and are only valid to evaluate different consumption patterns. When

Chapter 4 __________________________________________________________________________________________

36

possible, therefore, this chapter presented results in a relative manner.

Fig. 4.1. Three scale levels for food requirements and related land requirements. Requirements for the basic and subsistence levels are hypothetical requirements, for the cultural level actual requirements. On the basic level energy requirements are met by the consumption of wheat. It can only be maintained for a short period of time. Food requirements on the subsistence level are optimal from a nutritional point of view. They are based on a selected number of nutrient-dense foods, providing body’s health for the total life span. Food requirements on the cultural level form the actual consumption patterns and contain a broad variety of foods. They develop during generations. 4.3.2 Land requirements for basic and subsistence consumption On the basic level, energy requirements were met by the consumption of wheat. This menu only contained bread. For the calculation of related land requirements, the thesis used Dutch data on advised energy intake: 10 MJ (2400 kilocalories) per day per adult performing low physical activity (Voedingscentrum, 1998a). It based the assessment of land requirements on the subsistence level on Dutch recommendations for daily amounts of foods (see Table 4.1). In order to comply to energy requirements and nutritional constraints (55% of nutritional energy must be provided by carbohydrates, 35% by lipids, and 10% by proteins) the thesis replenished recommended amounts of foods by bulk food: wheat, potatoes, sugar, and fats. It obtained data from the Netherlands Nutritional Council (Voedingscentrum, 1998a). 4.3.3 Land requirements for culturally defined consumption patterns On the cultural level, the thesis assessed various consumption patterns and related land requirements. This was done for the Netherlands during the period 1950-1990 and for fourteen European countries and the United States in 1995. The thesis put food items into five categories: (1) beverages (beer, wine, coffee and tea); (2) fats (margarine, low fat spread, vegetable oil); (3) meat (beef and veal, pork, other meat, poultry); (4) dairy (full fat milk, semi-skimmed and skimmed milk, buttermilk, condensed milk, butter, cheese) and eggs; and (5) cereals, flour, sugar, potatoes, vegetables and fruits. Table 3.1 in Chapter 3 shows specific land requirements for these food items.

BASIC LEVEL Energy provided for by the

consumption of wheat

SUBSISTENCE LEVEL Energy and nutrients provided for

by the consumption of advised amounts of food items

Time span: total life

CULTURAL LEVEL Actual food consumption patterns

Time span: generations


_________________________________________________________________________________________

37

4.3.3.1 Inter-generational differences The chapter quantified changes in Dutch food consumption during the period 1950-1990 and assessed related land requirements. Data on Dutch consumption were available for twenty-seven consumption items. After World War II, the Netherlands Agricultural Economic Institute (LEI) and the Central Bureau for Statistics (CBS) have started to publish yearly data on per capita availability of foods in the Netherlands. Reliable information goes back as far as the year 1950. Earlier data were available - inventories started in 1933 - but were not compatible to more recent data. In order to calculate the influence of changing consumption patterns in the Netherlands, the thesis used the LEI/CBS data of 1980, 1985, and 1996. It obtained data on Dutch coffee and tea consumption from the Netherlands Coffee and Tea Council (Vereniging van Nederlandse koffiebranders en theepakkers, 1998). Table 4.1 Recommended daily amounts of food items per adult per day

Food item Recommended daily amounts (grams)

Bread 200 Potatoes 200 Vegetables 175 Fruits 200 Milk and milk products 375 Cheese 30 Meat (raw) 100 Meat products 23 Low fat spread 30 Margarine 15

Source: the Netherlands Nutritional Council (Voedingscentrum, 1998a) 4.3.3.2 Regional differences In order to assess regional variation of land requirements, the thesis performed calculations for consumption patterns of fourteen European countries and for the United States. For Europe, it based calculations on the availability of twenty commodities in 1995. The fourteen countries were the Netherlands, Belgium, Luxembourg, Denmark, Germany, Greece, Spain, France, Ireland, Italy, Austria, Portugal, Finland, and Great Britain. The thesis obtained data from various sources. Data on commodities came from the LEI/CBS (Landbouw-Economisch Instituut and Centraal Bureau voor de Statistiek, 1998); on coffee, tea, wine and beer consumption from the (Food and Agricultural Organisation of the United Nations (FAO), 1999); information on fat consumption from Eurostat (1993). To assess land requirements for the United States, it used data on food consumption as a basis for the calculations. These data derived from a consumer survey conducted by the US Department of Agriculture (Agricultural Research Service, 1999). However, calculations for Europe were based on the availability of foods, while for The United States data on consumption were used. There is a gap between available and consumed food caused by losses in the food chain, underreporting and non-food purposes (Gerbens-Leenes et al., 2001). In order to compare European and US land requirements, the thesis made a correction by calculating land requirements per recommended daily amount of energy. For the calculation, it applied Dutch data on advised energy intake. 4.3.4 Presentation of results Based on 1990 international data on food production, trade and consumption, the thesis calculated Dutch per capita land requirement for food in m2, square meters that are actually committed to food production, in the Netherlands as well as in countries trading with the Netherlands. The other results were also based on Dutch 1990 data, and must therefore be interpreted in a relative way. Where possible, the chapter presents results as land units, with Dutch per capita requirements in 1990 set to 100 land units, indicating the relative land requirement for the consumption pattern studied.

Chapter 4 __________________________________________________________________________________________

38

44..44 ►► RReessuullttss 4.4.1 Land requirements for the basic and subsistence levels Figure 4.2 shows land requirements for the basic and subsistence levels. Based on an average energy requirement of 10 MJ per capita per day and the consumption of wheat, the relative land requirement for the basic level was 23 land units. If recommended daily amounts of food advised by the Netherlands Nutritional Council were consumed, the relative requirement rose to 67 land units, three times more than the requirement for the basic level.

Fig. 4.2. Relative land requirements for the basic and the subsistence level, and actual relative land requirements for the cultural level. The latter requirements are based on existing food consumption patterns (see tekst for the explanation of land units). 4.4.2 Consumption and related land requirements for the cultural level 4.4.2.1 Dutch inter-generational differences Table 4.2 shows Dutch per capita food consumption between 1950 and 1990 and related energy intakes. In the category of beverages, the consumption of coffee, beer, and wine increased by a factor of eight to fifteen. Fat consumption did not change much, with only shifts among specific foods occurring. In the category of meat, total per capita consumption rose by a factor of three. Within this category, the thesis showed shifts among meat types. In 1950, people consumed mostly beef, veal, and pork, while consumption of poultry was negligible. By 1990, consumption of beef and veal had dropped in favor of poultry, while pork consumption had remained half of the total meat consumption. In the category of dairy and eggs, milk consumption decreased, while cheese consumption tripled. The thesis showed a shift towards milk varieties with lower fat content, however. In the category of cereals, sugar, potatoes, vegetables, and fruits, potato consumption dropped by 33%, while consumption of citrus fruits increased more than a factor of seven. Related energy intakes increased from 11.0 MJ per capita per day in 1950 to 15.2 MJ in 1980, but then dropped again to 12.7 MJ in 1990. Figure 4.3 shows that consumption changes had consequences for land requirements. Based on yields in 1990, Dutch land requirements for food increased from 72 land units in 1950 to 100 in 1990, an increase of 38%. Between 1950 and 1960, the rise was due to larger consumption of livestock products, fats, and beverages. Between 1960 and 1990, the rise was mainly caused by larger consumption of meat and beverages, while the consumption of fats, dairy, and eggs stabilized. The land requirement for the category of beverages tripled, mainly due to higher coffee consumption, which has a large specific land requirement, in combination with rising beer consumption. Between 1950 and 1960, the land requirement for the category of fats increased by 14%, but then remained stable. The

0

5 0

1 0 0

1 5 0

b a s ic le v e l s u b s is te n c e le v e l lo w e s t re q u ire m e n t c u ltu ra lle v e l

h ig h e s t re q u ire m e n t c u ltu ra lle v e l

rela

tive

land

requ

irem

ent i

n la

nd u

nits


_________________________________________________________________________________________

39

land requirement for the category of meat only doubled, despite a threefold increase of consumption. This relatively small increase was due to the shift from beef consumption, which has a relatively large specific land requirement, to poultry, which has a relatively small land requirement (see Table 3.1 in Chapter 3). In the category of dairy and eggs, the land requirement increased by 13% in the period 1950-1960, and then stabilized. Smaller milk consumption and a shift towards varieties with lower fat content and related smaller land requirements was compensated by larger cheese consumption, which has a large specific land requirement. The land requirement for the category of cereals, sugar, potatoes, vegetables, and fruits remained the same. In 1950, the land requirement for foods from livestock production systems was 44% of the total, in 1990 this contribution had increased to 47%, whereas the contribution of the category of beverages had increased from 4 to 12%. Table 4.2 Food consumption (kg per capita per year) and energy intake (MJ per capita per day) in the Netherlands during the period 1950-1990 Food item

Consumption (kg per capita per year)

1950 1960 1970 1980 1990 Beverages Beer 11 24 57 86 91 Wine 1 2 5 13 15 Coffee 1 4 6 7 8 Tea 1 1 1 1 1 Fats Margarine 17 20 18 13 10 Low fat spread 0 0 1 3 3 Vegetable oils 5 5 8 11 13 Meat Beef and veal 14 18 19 22 20 Pork 19 23 27 40 45 Other meat 2 2 3 3 3 Poultry 0 2 6 9 17 Dairy and eggs Full fat milk 188 127 107 60 42 Semi-skimmed milk 0 0 0 28 42 Skimmed milk 0 15 17 17 20 Buttermilk 16 14 11 10 11 Condensed milk 2 17 24 23 17 Butter 3 5 3 4 3 Cheese 5 8 9 14 15 Eggs 4 10 10 10 9 Cereals, sugar, potatoes, vegetables and fruits

Flour 81 71 57 54 66 Sugar 35 42 46 42 37 Potatoes 129 100 85 83 87 Vegetables 66 67 81 60 63 Fruits 30 37 34 37 34 Citrusfruits 8 17 24 39 61 Energy intake (MJ per capita per day)

11.0 11.7 11.7 15.2 12.7

Source: LEI/CBS 1980, 1985, 1996

Chapter 4 __________________________________________________________________________________________

40

Figure 4.3. The development of the relative per capita land requirement in the Netherlands during the period 1950-1990 based on 1990 yields for the five consumption categories: beverages; fats; meat; dairy and eggs; and cereals, sugar, potatoes, vegetables and fruits (c, s, p, v, f) (see tekst for the explanation of land units). 4.4.2.2 Regional differences Table 4.3 shows food consumption in the European Union in 1995. Large differences in the consumption of specific foods were apparent. In the category of beverages, beer consumption varied by a factor of six between Ireland and Italy, while the consumption of wine was largest in France and smallest in Ireland. Some countries had relatively large tea consumption, while in other countries, consumption was very small or even zero. In countries with the largest consumption of fats, people consumed twice as much fats than people in countries where consumption was relatively small. There was a variation of 62% between the highest and the lowest levels of meat consumption. The Scandinavian countries and Ireland showed large consumption of milk products, while consumption in Italy and Greece was relatively small. Butter consumption varied by a factor of sixteen, cheese consumption by a factor of four and a half, and cereal consumption by a factor of two. Some countries were large potato consumers (Ireland and Portugal), while consumption in countries, such as Italy and Finland, was small. Sugar consumption was large in Ireland; consumption of vegetables and fruits in Greece. Table 4.4 shows the relative per capita land requirements per country, per capita energy intakes (MJ year-1), and sums of the number of standard deviations per food item. The thesis divided these sums into two categories: food items with specific land requirements above and below the mean (6.6 m2 year kg-1). The mean European relative land requirement for food was 105, but large differences occurred between individual countries. Portugal showed the smallest land requirement (95), while the requirement for Denmark was largest (130), 37% larger than the lowest value. Energy intakes were 28% larger in Denmark than in the country with the smallest intake, the Netherlands. However, large land requirements were not necessarily correlated with large energy intakes. For example, the land requirement for Greece was only 1% larger than for Belgium, while Greek energy intakes were 5% larger. Relatively large land requirements were mainly caused by large consumption of foods with large specific land requirements. Table 4.4 shows that most of these consumption patterns were characterized by positive sums of the number of standard deviations of foods showing requirements above the mean of 6.6 m2 year kg-1 (e.g. Denmark and France). Medium land requirements were the result of medium consumption of foods with large specific land requirements, while foods with small land requirements were preferred (e.g. Ireland and the Netherlands).

0

10

20

30

40

50

60

70

80

90

100

1950 1960 1970 1980 1990

rela

tive

land

requ

irem

ent i

n la

nd u

nits

c,s,p,v,fdairy and eggsmeatfatsbeverages

Table 4.3 Per capita consumption in 1995 (kg year-1) in fourteen European countries, mean consumption and associated standard deviations

Consumption category Beveragesa Fatsb Meat c Dairy and eggsc Cereals, potatoes, sugar, vegetables and fruitsc

Beer Wine Coffee Tea Beef/ veal

Pork Other meat

Poultry Eggs Milk products

Condensed milk

Butter Cheese Cereals Potatoes Sugar Vegetables Fruits Citrus fruits

Danemark 121.5 25.0 8.6 0.4 43.6 17.6 64.2 1.2 15.3 15.9 141.7 0.0 9.6 15.9 74.6 57.1 40.5 80.0d. 49.0d. 15.0d.

France 34.5 63.2 5.2 0.2 e. 28.1 35.9 5.3 22.6 16.0 101.6 0.7 8.3 23.3 76.2 58.5 33.3 124.0d. 58.0d. 24.0d.

Greece 36.0 14.9 2.2 0.0 e. 19.6 24.8 13.6 17.7 10.6 64.0 0.0 1.2 23.4 138.5 87.1 25.6 308.1 80.1 43.3

Italy 23.3 57.5 4.8 0.1 30.9 25.9 33.1 1.7 18.4 10.5 68.6 0.6 2.6 19.0 118.1 38.3 25.6 174.6 68.2 39.7

Austria 113.1 32.0 6.0 0.2 e. 19.6 56.9 1.2 15.3 13.8 98.9 2.4 5.0 14.2 67.8 60.5 39.7 79.8 77.9 17.2

Belgium/ Luxemburg

109.1 21.0 2.0 0.1 32.5 21.2 46.6 2.1 23.1 14.5 83.1 2.1 5.9 14.2 72.7 94.1 42.4 99.5 68.7 32.3

Ireland 137.0 4.4 1.8 3.1 e. 14.5 37.9 7.2 30.9 9.2 176.3 0.3 3.6 5.3 80.7 173.8 43.0 90.7 33.8 15.2

Netherlands 86.5 11.9 9.0 1.0 24.9 19.8 46.3 1.3 20.1 15.3 129.6 7.1 4.0 14.1 58.4 87.6 32.7 93.8 64.1 45.4

Spain 58.3 36.5 3.8 0.1 30.2 12.7 55.3 6.6 25.5 15.3 133.7 1.0 0.6 7.1 72.1 86.3 31.7 153.9 64.6 46.3

Sweden 66.0 12.4 9.4 0.3 e. 18.2 36.1 0.7 7.9 12.0 150.7 1.2 5.5 15.6 65.6 57.7 40.3 98.9 41.6 1.2

Finland 88.3 5.6 7.8 0.2 e. 19.1 32.2 0.5 8.8 11.8 199.7 0.0 5.4 13.5 69.2 56.9 33.3 76.9 30.5 15.5

Germany 131.1 23.4 7.1 0.2 21.2 16.6 55.0 1.1 13.4 13.8 91.1 5.4 7.2 18.4 75.1 72.9 32.0 86.2 64.7 29.0

Great Britain

102.0 11.4 2.2 2.3 30.0 17.5 23.1 6.0 25.1 10.1 131.2 2.5 3.3 7.8 81.6 101.3 36.8 99.4 39.5 17.3

Portugal 67.4 58.4 3.3 0.0 e. 17.6 34.7 3.6 23.0 8.6 100.5 0.4 1.5 7.2 82.4 138.7 29.2 123.7 77.9 25.5

Average ± sd.

83.9 ± 5.8

27.0 ± 9.3

5.2 ± 2.7

0.6 ±

0.9

30.0 ± 6.5

19.1 ± 3.9

41.6 ±2.2

3.7 ± 3.6

19.1 ± 6.3

12.7 ± 2.5

119.3 ± 38.2

1.7 ± 2.1

4.6 ± 2.6

14.2 ± 5.6

80.9 ± 20.7

83.6 ± 34.9

34.7 ± 5.6

120.7 ± 58.9

58..5 ± 6.2

26.2 ± 3.2

a. Source: FAO (1999); b. Source: Eurostat (1993); c. Source: LEI-DLO/CBS (1998); d. Data of 1992; e. Mean EU fat consumption (30 kg)

Chapter 4 __________________________________________________________________________________________

42

Relatively small land requirements were mainly found in countries with a relatively small consumption in all categories (e.g. Great Britain and Portugal). Sums of numbers of standard deviations were negative. Table 4.4 Relative land requirements in land units for food consumption patterns in fourteen European countries in 1995, related energy intakes (MJ per capita per year) and sums of number of standard deviations for foods with relatively large or small specific land requirements Consumption pattern of

Land requirement (land unitsa.)

Energy intake (MJ per capita per year)

Sum of number of deviations per consumed food item

land requirement > 6.6 m2 year kg-1

land requirement < 6.6 m2 year kg-1

Denmark

130 5991 +4.3 +2.3

France

118 5241 +5.3 0.0

Greece

107 5607 +1.4 +0.5

Italy

107 5202 +0.5 +1.0

Austria

106 5104 -0.5 +2.8

Belgium/ Luxemburg

106 5353 -0.8 +4.1

Ireland

104 5491 -0.7 +2.5

Netherlands

102 4672 0.0 +3.8

Spain

101 5028 -4.7 +4.7

Sweden

100 4698 +1.0 -7.0

Finland

99 4666 -0.1 -7.6

Germany

98 4851 -0.6 +1.7

Great Britain

96 4855 -0.6 -2.5

Portugal

95 5008 -4.3 +0.3

a. See text for the explanation of land units. In three countries, France, Italy and Ireland, a correlation between beer and wine consumption occurred in which large consumption (mean + SD) of one beverage correlated with small consumption (mean - SD) of the other. In Ireland and Great Britain, this effect was also observed for the large tea and small coffee consumption. For Greece and Ireland, the chapter showed a relationship between large consumption of milk products and small cheese consumption. In the category of meat, large consumption of a specific type in combination with small consumption of another occurred in Greece, Ireland, and Spain. Other correlations between the consumption of specific foods were not evident. Figure 4.2 shows hypothetical land requirements and estimates of actual land requirements. In order to make an estimate for the range of land requirements related to food consumption patterns, the chapter used European patterns as the basis for the calculations. For the assessment, it summed lowest and highest values for the five consumption categories. It termed the lowest outcome the 'lowest cultural level', and the highest outcome the 'highest cultural level'. Relative land requirements varied between 23 and 143, a difference of a factor of six. The estimate of relative land requirements based on European consumption patterns in 1995 varied between 72


_________________________________________________________________________________________

43

and 143, a difference of a factor of two. All requirements based on existing patterns were larger than the requirement for the subsistence level. 4.4.2.3 Correction for energy intake Figure 4.4 shows relative land requirements per advised energy intake for the European countries and for the United States. Large differences were found among the European countries on the one hand, as well as between Europe and the United States on the other. European relative land requirements varied between 69 land units (Ireland) and 82 land units (France) per 10 MJ, a difference of 19%. The relative requirement for the United States was 100 land units per 10 MJ, 34% higher than for the European mean. This was mainly due to the high proportion needed for meat in the US: 45% of the total. In Europe, the country with the largest requirement for meat (Spain) showed a proportion of only 32%. The second large difference related to U.S. and European consumption patterns was the relative low requirement for dairy in the U.S., only 7% of the total. In Europe, the country where dairy consumption was smallest (Portugal) still showed a requirement that was almost twice as large as for the U.S.

Fig. 4.4. Relative, per capita land requirements for food in 1995 based on an energy intake of 10 MJ per day, divided over five consumption categories for fourteen European countries and for the United States (see tekst for the explanation of land units). c, p, s, v, f = cereals, potatoes, sugar, vegetables and fruits. 44..55 ►► DDiissccuussssiioonn Cultural, non-physiological requirements claim a substantial part of the land area needed for food. Therefore, in western countries, the influence of food consumption patterns on related land requirements is substantial, resulting in large regional as well as inter-generational differences. The physiological requirement is only 67 land units. In Europe in 1995, the non-physiological requirement varied between 28 land units for the Portugese consumption pattern and 63 land units for the Danish pattern, while in the Netherlands the non-physiological requirement rose from 5 land units in 1950 to 33 in 1990. In Europe, regional and inter-generational differences are mainly caused by variation in the consumption of meat and different drinking habits. Affluence seems to be related to high fat consumption, claiming about one third of the total agricultural area. For example, in the Netherlands after World War II, increasing incomes caused a rise in fat consumption due to larger consumption of affluent, fatty foods such as ice cream, cakes, and high-fat junk foods. Severe negative health effects (De Wijn, 1968) were the reason that industry developed low-fat foods so that total fat consumption did not rise further.

0

20

40

60

80

100

Irelan

d

Portugal

Greece

Belgium/Luxe

mburg

United K

ingdomSpain

German

yIta

ly

Austria

Finland

Sweden

Danem

ark

Netherla

nds

France

United S

tates

rela

tive

land

requ

irem

ent i

n la

nd u

nits c,p,s,v,f

dairymeatfatsbeverages

Chapter 4 __________________________________________________________________________________________

44

The attitude of agricultural scientists towards the issue of land requirements has assumed that these requirements are mainly determined by the agricultural system applied, while food consumption is simplified. Penning de Vries et al. (1995) and Bouma et al. (1998) have recognized only three diets, a vegetarian one, a moderate one, and an affluent one with meat, the last requiring three times as much biomass as the first. This paper demonstrated that not only meat consumption, but also the consumption of fats and beverages require large agricultural land areas. Food security studies have not assessed the influence of beverages on land requirements because they are not considered to be food. However, in Europe, the production of only four beverages - beer, wine, coffee and tea – required 10% of the total agricultural land area. The amount of land needed for beverages is actually even larger because no data were available for the consumption of soft drinks and juices. Consumption of these beverages has risen in the past decades; Table 4.2 shows a large increase in the consumption of citrus fruits, a basic ingredient for juices, and a rise in the consumption of sugar, an ingredient for soft drinks. There are several reasons why requirements for beverages should be included in the calculations. The first reason is that ingredients for beverages are produced in agricultural systems, hence they stake a claim on the same agricultural resources as foods. Sometimes even the same ingredients are used. For example, 28 kg of barley are required for the production of 100 liters of beer (Kramer and Moll, 1995), while 400 kg of barley is needed for the production of 100 kg of pork (Nonhebel, 2001). In Great Britain, per capita consumption of beer is 100 liters per year and of pork 23 kg, requiring 28 and 100 kg of barley respectively. This example shows that the claim of beverages on agricultural resources cannot be ignored. The second reason is that, when compared to foods, the nutritional value of many beverages is small, implying that, in contrast to food, there is no clear physiological limit for beverage consumption. There seems to be a maximum to the physical consumption of certain foods, such as, for example, meat. Jobse-van Putten (1995) has shown that Belgian and Dutch households belonging to the upper-classes consume less meat than lower-class households. However, if consumption is regarded as expenditure on food, the upper-classes prefer more expensive types of meat, such as veal and lamb, while the lower classes buy the cheaper pork. A second example comes from Canada. In the Dene/Metis communities of indigenous peoples in the western Canadian Arctic, increasing affluence has caused a shift in traditional diets, mainly based on animal foods harvested from the local environment, to market foods containing more items of vegetable origin, such as sugar (Receveur et al., 1997). It can therefore be expected that if affluence increases further in western countries, meat consumption will stabilize but beverage consumption will probably rise accordingly, generating land claims. In the near future, consumption patterns will form a very important variable for total land requirements on a global scale, especially dietary changes in direction towards the larger consumption of beverages, fats, and foods of animal origin. In this respect, differences between the U.S. and European food consumption patterns are large. If Europeans shift towards the American level of meat consumption patterns, land requirements will rise by 17%. On the other hand, if Americans adopt European dairy consumption, the U.S. land requirement for food will rises by 12%. On a global scale, it should be realized that a large part of the population is undernourished. Vitamin A deficiencies in particular, which can be solved by larger fat consumption, are a major a problem (Whitney and Rolfes, 1999). The provision of enough quality food for the total world population on at least the subsistence level requires an increase in the agricultural production of certain commodities. In developing countries, rising incomes not only change food choice but also increase the average, total, per capita food supply. This change in direction has been demonstrated for Benin, Bhutan, and Costa Rica (Van Vuuren and Smeets, 2000). If consumption patterns in developing countries shift towards the affluent menus of western countries, related per capita land requirements will rise substantially. In these regions, two or even three-fold increase in requirements, which is much larger than the influence of the growth in the world population, is certainly possible. Shifts towards more affluent diets will concur with other claims on the available land, such as infrastructural developments, ecological forms of agriculture, biodiversity, energy crops, or food production for the world market. 44..66 ►► CCoonncclluussiioonnss Absolute increase of food supply in third world countries on the one hand, and dietary shifts in developing as well as in developed countries on the other cause larger claims on agricultural land resources. In developing countries, requirements will first need to fulfill basic dietary needs, provided for by nutrient-rich foods such as fats, foods of animal origin, and fruits. Food security studies have shown that future generations can be fed in a way that meets nutritional constraints. However, it should be realized that the social and cultural requirements of food claim large parts of available land


_________________________________________________________________________________________

45

resources. Social and cultural claims of food on land requirements are substantial and will possibly increase in the near future. The amounts of specific foods that form the basis of consumption patterns can change rapidly. As has been shown for the Netherlands, dietary changes can take place in a relatively short period of time. In western countries, the increasing consumption of beverages, claiming the same agricultural areas as foods, is important. In order to define future land requirements for food on a global scale, studies should incorporate social and cultural needs of food consumption patterns into the calculations. These requirements have a large impact and can change rapidly. Since it is not certain that yields will rise further in the future, trends towards the consumption of foods associated with affluent life styles will bring with them a need for more land, since requirements are relatively large. This dietary change in direction is especially important for developing countries, where many people are still undernourished. The effects of changes in food consumption patterns on land requirements will be even larger than the growth in the world population. This effect might double the need for agricultural land.

46

47

CChhaapptteerr 55

CCrriittiiccaall wwaatteerr rreeqquuiirreemmeennttss ffoorr ffoooodd,, mmeetthhooddoollooggyy aanndd ppoolliiccyy

ccoonnsseeqquueenncceess ffoorr ffoooodd sseeccuurriittyy∗∗

Abstract Food security and increasing water scarcity have a dominant place on the food policy agenda. Food

security requires sufficient water of adequate quality because water is a prerequisite for plant growth. Nowadays,

agriculture accounts for 70% of the worldwide human fresh water use. The expected increase of global food

demand requires a great deal of effort to supply sufficient fresh water. If a doubling of agricultural production goes

along with a doubling of the use of water, current fresh water resources are probably not sufficient in the long run.

The objective of this chapter is to develop a generally applicable method for the assessment of crop growth-

related water flows or ‘transpirational’ water requirements of agricultural crops. Traditionally, agricultural studies

have made assessments of water requirements for specific situations to provide a yield. This chapter uses the

agricultural information the other way around. Water had to be present for a growth to occur. Based on the strong

linearity of processes taking place in all green plants, the chapter develops a method to calculate the growth-

related factor of crop water requirements, assesses the impact of crop characteristics on water requirements, and

evaluates options to reduce the use of water by changing food consumption patterns. The chapter calculates

‘transpirational’ water requirements for a representative group of crops with different functions for human

nutrition, such as staple crops, vegetables, and livestock fodder. For C3 crops in a temperate climate, 0.11 liters

are needed to produce 1 gram of glucose; for C4 crops in a tropical climate, 0.09 liters. Water requirements per

unit of dry mass differ by a factor of two. These differences arise from differences among harvest indices of crops

and their chemical composition. Differences in requirements per unit of nutritional energy, however, are low.

Therefore, options to reduce the use of water by qualitative changes of food consumption patterns are few.

The chapter distinguished between site-specific and crop growth specific water flows. In this way, it quantified a

central flow of the hydrological system, the water flow that actually passes a crop and is directly related to the

photosynthesis process. If yields increase, this water demand increases with the same factor. The results show

critical water requirements in agriculture. However, these results are intended to improve the insight into

hydrological systems and must always be used in combination with locally, variable water needs. The results

have two important consequences for food policy issues. Firstly, the chapter shows only small differences in

water requirements among crops. Secondly, results indicate that under rainfed conditions, a doubling of food

production on existing land areas does not imply a doubling of water use but only of ‘transpirational’ water use.

This flow forms a minor part of total water requirements in agriculture.

55..11 ►► IInnttrroodduuccttiioonn Food security for a growing world population requires the availability of sufficient water of adequate quality. Increasing water scarcity and the need for more food have a dominant place on the food policy agenda (Pinstrup-Andersen, 2000). Today, irrigated agriculture requires 70% of total human ∗ This chapter is a slightly adapted version of Gerbens-Leenes, P.W., Nonhebel, S., 2004. Critical water requirements for food, methodology and policy consequences for food security. Food Policy, 29, 547-564.

Chapter 5 __________________________________________________________________________________________

48

fresh water use (FAO, 2003a; Falkenmark, 1989b; Rosegrant and Ringler, 1998; Rockstrom, 1999). In some parts of the world, such as Eastern Asia and the former USSR, water is already a scarce resource (Penning de Vries et al., 1995). In these parts of the world, the use of water for agricultural purposes has to compete with other uses such as urban supply and industrial activities (Falkenmark, 1989). Moreover, projections for 2025 indicate that more than half of the world’s population will live in regions dependent on food imports due to water scarcity (Falkenmark, 1997). Additionally, recent research has suggested that climate change will lead to major shifts in the spatial and temporal patterns of precipitation (IPCC, 2001). In Southern Europe, for example, long-term projections (2070s) indicate a decrease of water availability between 25 and 50%, while in large parts of France, an important producer of wheat, decreases lie between 10 and 25% (Lehner et al., 2001). So far, the growth of global agriculture has been sufficient to meet the growth of demand (FAO, 2003a). By 2050, the United Nations’ medium projection estimates global population to be 50% larger than in 2003 (United Nations Population Division, 2002), while FAO projections (FAO, 2003a) indicate that per capita food demand increases by 9% in the next thirty years. Again, the resulting increase in demand requires large efforts from agriculture (Tilman et al., 2002) and implies huge challenges on the availability of sufficient water resources. The required growth of global food production can be achieved via three routes, an increase of agricultural land areas, an increase of yield levels per unit of land, or an increase in cropping intensities (i.e. increasing multiple cropping and shorter fallow periods). If agricultural land areas are increased, water inputs will probably increase with the same factor because input per unit of land (m3 per m2) remains the same. This large increase of water inputs in agriculture is probably not possible, implying that water scarcity is the reason that the provision of food for a growing world population must focus on a more efficient use of water per unit of crop (m3 per kg) (Wallace, 2000). This requires a fundamental insight into the relationship between yield levels, crop types and related water requirements. Traditionally, agricultural research has assessed local water needs. This has resulted in large variation on the amount of water to produce a unit of food, even for equal food types. To grow 1 kilogram of potatoes, for example, Renault and Wallender (1999) calculate a requirement of 105 liters of water, Pimentel and Houser (1997) 500 liters of water, and Wallace (2000) 840 liters of water. Variation of results is caused by variation among local hydrological systems and implies that information on water use obtained at a specific site cannot be used for other locations. General conclusions on surplus water requirements related to an increase of food production or differences in water needs among foods, therefore, must be obtained through another methodology. An important characteristic of the methodology should be that it does not take locally determined, highly variable water flows into account but focuses on similar processes in biomass production. The objective of this chapter is to develop a methodology for the quantification of the amount of water that had been required to produce a unit of food. Based on the presence of a harvest, it argues that water had been available for growth to occur. This is the ’transpirational’ water flow summed over the complete growing season. This flow forms the absolute minimum or critical water requirement to provide a yield. In this way, the chapter provides information on the impact of crop characteristics on water requirements and on surplus requirements related to an increase of food production. The results provide a better understanding of the consequences of changes in food production systems on water requirements, and indicate the direction and magnitude of changes in water requirements. By improving the insight into hydrological systems, the thesis contributes to the information need relevant for the water scarcity and food security issue. 55..22 ►► SSyysstteemm ddeessccrriippttiioonn Agricultural and hydrological systems are very complex. For the assessment of the amount of water needed for the production of a unit of food, this chapter strongly simplifies these systems. Firstly, it simplifies the agricultural system by defining seven hypothetical crops representing the globally most important crops for human nutrition. Secondly, it neglects the locally determined water flows and focuses on the amount of water that has actually passed the biomass present after harvest, the ’transpirational’ water flow. 5.2.1 Food crops and their place in human nutrition Globally, a large variety of food items is available for human consumption. Food items are products of animal or vegetable origin that form the end of a food production chain, often requiring several commodities deriving from primary or secondary production systems. Primary production grows crops,

Critical water requirements for food, methodology and policy consequences for food security

__________________________________________________________________________________________

49

secondary production uses these crops for livestock products. Although the variety of foods is large, globally there is only a limited amount of crops that form the basis for human food consumption patterns. The fifteen most important arable crops or categories of crops in order of decreasing annual production are: sugar cane, root crops, vegetables, maize, paddy rice, wheat, fruits, potato, sugar beet, cassava, soybean, barley, pulses, oil seed rape, and sorghum (FAO, 1999). The chemical composition of an arable crop determines its function for human nutrition. Staple crops, such as wheat and potatoes, are consumed for their carbohydrate content. Other crops are rich in proteins, such as pulses, or provide oil, such as oil seed rape. Grass is an important non-arable crop because it is significant for livestock fodder and therefore for human nutrition. Chapters 3 and 4 have shown that there are large differences in resource use among specific food items and food consumption patterns. Differences in the use of land resources, for example between potatoes and cereals, are already known from agricultural history. In Western Europe in the 18th century, potatoes were introduced on a large scale to prevent hunger because potatoes provided two to three times as much nutritional energy per unit of land area as cereals (Jobse-van Putten, 1995). An important question is whether changes in the use of crops for food cause changes in the use of water. 5.2.2 Crop production The basis for primary production is the photosynthesis process in which solar photonic energy is converted into the chemical energy of glucose. The radiation on the ground surface available for photosynthesis is termed global radiation and is expressed in megajoules per unit of land area per day (MJ m-2 day-1). The efficiency of photosynthesis shows large variation, but on the scale of a complete growing season and under conditions without shortage of water or nutrients, a linear relationship between intercepted global radiation and above ground biomass is observed (e.g. Goudriaan et al., 2001; Monteith, 1977a). This relationship exists for all plants, from agricultural crops to trees. In the Netherlands, for example, crops rich in carbohydrates, such as wheat, potato, sugar beet, and maize, grown under near-optimal growth conditions, form dry matter at a rate of 200 kg per hectare per day (Sibma, 1968). Although the formation of glucose is the basis for all dry matter production, variation among crops is large and results in different functions of specific crops for human nutrition. Four important characteristics of crops are responsible for this variety: the chemical composition, the harvest index, the dry matter content, and the biochemical pathway of photosynthesis. The chemical composition of a crop determines the amount of dry matter formed per unit of radiation. Groups of organic compounds are, for example, lipids, lignin, proteins, carbohydrates, and organic acids. Although the synthesis of glucose per MJ of intercepted global radiation is a uniform process for all plants, the amounts of glucose required for a unit of dry mass differ among plants and depend on their chemical composition. For fats, the amount is relatively large, for carbohydrates relatively low. Penning de Vries (1983) has calculated the conversion factors (CVFs in grams product per gram glucose) with great accuracy. The value for fats is 0.31, for proteins 0.52, and for complex carbohydrates 0.78. In general, agriculture grows crops for their reproductive or storage organs that have an economic value. It grows cereals, for example, to produce grains and potatoes to produce tubers. Agriculture grows other crops, such as vegetables, for their leaves or stems. The growth of these organs, however, requires the preceding growth of complete plants with stems and foliage. Figure 5.1 shows the harvest index, the ratio of the economic yield and the total biomass production, that determines the fraction of a yield available for human consumption. The difference between the total biomass production and the economic yield is the rest fraction. A low harvest index implies that crops make large glucose investments in this fraction of non-edible biomass and have a small economic yield. Data on harvest indices are available from literature. The dry matter contents of economic yields and rest fractions show large variation among crops. Potatoes, for example, have a dry matter content of 25%, while wheat has a dry matter content of 85% (Goudriaan et al., 2001). The third crop characteristic, the dry matter content, determines the amount of fresh biomass produced per unit of radiation. The biochemical pathway of photosynthesis takes place via two different routes, which are characterized by the length of the C-skeleton of the first stable product, comprising either three or four carbon atoms. Since these pathways are species specific, typical C3 and C4 plants are distinguished. Potato and wheat, for example, are C3 species and maize is a C4 species. C4 plants have a higher optimum temperature for photosynthesis than C3 plants. Under optimal conditions, the efficiency of solar energy conversion is 40% higher for C4 than for C3 species (Monteith, 1977b).

Chapter 5 __________________________________________________________________________________________

50

Fig. 5.1. The two components of the total biomass production, the economic yield and the rest fraction. Crops show a biological yield, i.e. the total biomass produced. This yield consists of an economic yield: that part of the total yield that can be traded on the food commodity market and is suitable for human consumption, and a non-food part or rest flow that has little economic value and is not edible for humans. The ratio of the economic yield and the total biological yield is termed the harvest index and depends on crop characteristics. 5.2.3 Water flows at a crop field Crop growth requires water in the root zone of plants. The hydrological system of a crop field comprises six main water flows, precipitation, drainage, run-off, evaporation from the soil, transpiration from the crop leaves, and irrigation. These flows interact and can therefore not be assessed independently. Moreover, local hydrological systems are highly variable in space and time, and sensitive to land use and its change (Schulze, 2000). In the Netherlands, for example, rainfall varies between 400 and 1200 mm per year (Buishand and Velds, 1980). At a field level, a water balance approach distinguishes between vertical and horizontal components of root zone water flows. Figure 5.2 shows a simplified overview of the two stocks, the crop root zone and the above ground crop mass distinguished at the field level. Flow 1 represents the total supply composed of precipitation (vertical) and exogenous inflow (horizontal). Flow 2 represents the horizontal output of water lost to rivers and aquifers, and the vertical downward flow leaving the root zone of plants to lower layers (groundwater), and eventually to open water. Flow 3 is the water evaporated from the soil, and flow 4 represents the water flow that actually passes through crops and is related to plant growth. Transpiration and evaporation from the soil surface are termed evapotranspiration (Buishand and Velds, 1980; FAO, 1992). Transpiration efficiency has been widely investigated at the leaf and the plant levels, but rarely of a crop at the field level owing to the difficulty in directly measuring transpiration or in separating transpiration from evapotranspiration (Zhang et al., 1998). Under rainfed conditions, agriculture grows crops using only natural water flows. The availability of water is sometimes a limiting factor for crop growth, however, and agriculture applies irrigation for higher yields. This thesis distinguished variable, site-specific water flows (flow 1, 2 and 3), and a crop growth-related flow (flow 4). The size of the site-specific flows is determined by local factors and expressed in liters per unit of land area; the size of flow 4 (i.e. transpiration) is determined by crop growth and expressed in liters per unit of mass. The thesis terms this flow passing crops ‘transpirational’ water. Evapotranspiration, therefore, consists of a variable, site-specific water flow and a growth-related flow. Solar radiation is the principal driving force for the evaporation of water but also for the photosynthesis process. If water is limited, plant growth is limited with the same ratio. Agriculture uses this relationship to assess water requirements during crop growth. The relationship between growth and water requirements can also be applied the other way around, however. There are many equations available to estimate potential evapotranspiration requiring the input of meteorological data (Smith et al., 1991). The FAO (1992), for example, used such an equation to develop the computer program CROPWAT, a useful tool for farmers for irrigation planning and management.

Economic Yield Rest Fraction

Total Biomass


__________________________________________________________________________________________

51

55..33 ►► MMaatteerriiaallss aanndd mmeetthhooddss 5.3.1 Starting points Traditionally, agricultural studies have assessed water requirements for specific situations to provide a yield. This thesis started from the other side, the availability of a yield, and related the presence of biomass to the availability of water for plant growth. It argued that water had to be present for a growth to occur. Based on the strong linearity of processes taking place in all green plants, the chapter developed a method to calculate the growth-related factor of crop water requirements and assessed the impact of crop characteristics on these requirements. By assessing water losses related to photosynthesis, it was possible to quantify one of the flows of the hydrological system in a generally applicable way. Fig. 5.2. Simplified overview of the two water stocks, the crop root zone, and the crop mass in a crop field. Flow 1 represents the total supply composed of precipitation (vertical) and exogenous inflow (horizontal). Flow 2 represents the horizontal output of water lost to aquifers and rivers, and the vertical downward flow leaving the root zone of plants to lower layers (groundwater and eventually to open water). Flow 3 is the water evaporated from the soil, and flow 4 represents the water flow that actually passes through crops and is related to plant growth. Flow 3 and flow 4 are termed evapotranspiration. From physics, a linear relationship between the evaporation of water and radiation is known, while agricultural research has shown a linear relationship between the synthesis of dry matter and radiation (e.g. Goudriaan, 1982; Monteith, 1977a; Goudriaan et al., 2001). The gain in dry weight per unit of water loss has been described for the first time by Woodward (1699) and has been confirmed by the standard reference of De Wit (1958). If water is limited, photosynthesis is limited with the same ratio. This implies that one of the water flows of the hydrological system of a crop field, the ’transpirational’ water flow that actually passes a crop, is linearly related to yield levels. Within carefully known conditions, this thesis calculated the amount of water transpired from the amount of above ground biomass using equations from meteorology and agricultural science. This theoretical approach made it possible to separate transpiration from evapotranspiration and to assess the ’transpirational’ water flow by combining data on dry matter produced per unit of radiation, information on CVFs of crop organic matter, and data on transpiration. Data on the forming of dry matter (grams per MJ) and CVFs (grams product per gram glucose) were available from agricultural studies. Data on transpiration were calculated using a model adopted from meteorological studies requiring input of climate data.

Open Water Ground Water

Aquifers

Crop Field

1

2

4

3

4

Crop Mass

System Boundary

Atmosphere

Crop Root Zone

Chapter 5 __________________________________________________________________________________________

52

5.3.2 Hypothetical crops as representatives for crop types The thesis calculated ‘transpirational’ water requirements for a representative group of crops with different functions for human nutrition, such as staple crops, vegetables, and livestock fodder. The four crop characteristics of the sixteen globally most important crops show variation. Based on this variety, the thesis distinguished ten categories of crops. These categories show: (i) a low harvest index, (ii) a high harvest index, (iii) a low water content, (iv) a high water content, (v) a relatively low content of carbohydrates, (vi) a relatively high content of carbohydrates, (vii) a relatively low content of proteins, (viii) a relatively high content of proteins, (ix) a relatively low content of fats, (x) a relatively high content of fats. C3 and C4 crops are found in every category. In general, biological systems show natural variation. Even for a specific crop type, the chemical composition of the dry matter, the harvest index, and the water content vary. Habekotté (1996), for example, has demonstrated that carbohydrate, protein, and oil contents of the seeds of winter oil seed rape are variable. Since these macronutrients all have different CVFs, natural variability results in varying glucose requirements per unit of weight. For the provision of general information on ’transpirational’ water requirements, natural variation forms a complication. This chapter, therefore, defined seven hypothetical crops (H-crops) that were considered representative of the ten crop categories and were derived from existing crops. These were: potato (Solanum tuberosum L.), wheat (Tricicum aestivum L.), maize (Zea mays L.), field pea (Pisum sativum L.), winter oil seed rape (Brassica napus L.), grass (Gramineae), and spinach (Spinacia oleraea L.). The chapter added H-maize as a representative of C4 crops, spinach as a representative of vegetables, and H-grass for its importance as a livestock fodder. Table 5.1 shows the main characteristics of the H-crops. 5.3.3 Radiation use efficiency and glucose for growth Radiation use efficiency (RUE) is the conversion factor between the amount of radiation intercepted or absorbed by a plant canopy and the amount of carbon dioxide fixed or biomass produced (Arkebauer et al., 1994; Monteith, 1994). This thesis considered global radiation, the flux of solar radiation reaching the surface of the earth (MJ m-2 day-1). The RUE is often based on field studies assessing the relationship between the formation of dry matter and global radiation (Monteith, 1994). This thesis defined the RUE as the amount of glucose produced per unit of global radiation (grams MJ-1) that was calculated using data of field studies of Monteith (1977a) and the CVF of carbohydrates from Penning de Vries (1983). Based on the chemical composition of the attained yields of the H-crops, the thesis used the RUE and CVFs to calculate the amount of glucose required for producing a unit of mass of the economic yield and of the rest fraction. Based on data of harvest indices, the thesis allocated the amount of glucose required to produce the rest fraction to the economic yield. 5.3.4 The transpiration of water For the calculation of evaporation, several estimation methods have been developed requiring meteorological data (Smith et al., 1991). A simple equation adopted here is a modified version of the Makkink equation (KNMI and CHO, 1988). It uses global radiation and the slope of the saturation vapor temperature curve as input data. Transpiration was calculated by:

T (K) = 0.65 λγK

ss+

in which:

T = transpiration of a crop (liters m-2 = mm)

s = slope of the saturation water vapor temperature curve (hPa oC-1)

γ = psychometric constant = 0,67 hPa oC-1 (Buishand and Velds, 1980)

λ = latent heat of vaporization of water = 2.26 MJ kg-1 (Verkerk et al., 1986)

K = global radiation (MJ m-2)


__________________________________________________________________________________________

53

In this equation, transpiration mainly depends on global radiation because the sensitivity of T to air

temperature is small. Between 15 and 35 oC, the term γ+s

s varies between 0.6 and 0.8. Assuming

air temperatures in the growing season between 15 and 25 oC, the thesis took an average value of 0.7 for crops grown in a temperate climate. For crops grown in a tropical climate, it used the value of 0.8. 5.3.5 ’Transpirational’ water requirements Without limiting factors other than water, transpiration losses and the ability of roots to take up water determine final yields. For the calculation of total transpiration related to an attained yield, this chapter used a step-by-step approach. Information on the amount of glucose produced per unit of global radiation was combined with data on transpiration. This provided data on amounts of glucose produced per unit of water. Next, it assessed the amount of glucose required for producing an attained yield. Finally, the chapter calculated the amount of water required to produce this yield. 5.3 shows the flow chart, the six calculation steps, and inputs for the calculation. Based on data of carbohydrate production of C3 crops per unit of radiation (1.4 grams per MJ global radiation) that was derived from field studies of Monteith (1977a) and the CVF of carbohydrates (Penning de Vries, 1983), Step 1 calculated the amount of glucose per MJ global radiation available for the growth of C3 crops. For C4 crops grown in a tropical climate, the thesis assumed a 40% increase in efficiency (Monteith, 1977b). The thesis calculated transpiration by using the Makkink equation. Step 2 assessed the amount of water transpired (liters per MJ global radiation). Step 3 combined results of step 1 and 2, and calculated the amount of water needed to produce glucose available for crop growth (liters per g). Based on the harvest index, Step 4 calculated the total amount of biomass required to produce a crop. Based on the chemical composition of the H-crops as shown in Table 5.1 and data on CVFs (Penning de Vries, 1983), Step 5 assessed the amount of glucose required for the total biological yield (grams per gram). Finally, Step 6 assessed the ’transpirational’ water requirement for the crop (liters per kg) by combining results of step 3 and 5. The thesis calculated the ’transpirational’ water requirement for maize for two climatic regions, a temperate climate with maximum temperatures between 15 and 25 oC, and a tropical climate with temperatures between 30 and 40o C. 55..44 ►► RReessuullttss The first three steps shown in Figure 5.3 provided general information, step 4 the ’transpirational’ water requirements of crops. Step 1 provided the amount of glucose available for crop growth: for C3 crops, 1.8 grams glucose per MJ global radiation; for C4 crops, the amount is larger: 2.5 grams per MJ. Step 2 provided the amount of water transpired. Potential crop transpiration is 0.20 liters per MJ

global radiation in temperate climates and 0.23 liters per MJ in tropical climates. Step 3 provided water requirements of glucose production. For C3 crops in a temperate climate, 0.11 liters are needed to produce 1 gram of glucose; for C4 crops in a tropical climate, 0.09 liters. Table 5.2 shows the ‘transpirational water’ requirements for the seven H-crops provided by step 4 in liters per kilogram economic yield, liters per kilogram dry mass and liters per kJ nutritional energy. For H-maize, Table 5.2 shows results for two climatic conditions, a temperate and a tropical climate. Differences among ’transpirational’ water requirements for crops depend on specific glucose requirements for crop biomass determined by crop characteristics such as biomass composition and harvest indices. Crops showing relatively large water contents, for example, use little glucose for biomass production resulting in low ’transpirational’ water requirements. Table 5.2 demonstrates that the large water contents of H-grass, H-spinach (92%), and H-potato (75%) result in low ’transpirational’ water requirements in liters per kilogram product. In general, grains have small water contents and therefore relatively large ’transpirational’ water requirements. The requirement for H-wheat, for example, is a factor of fifteen larger than for H-grass. Proteins and fats require more glucose than complex carbohydrates and thus more water. The large oil content of H-winter oil seed rape, therefore, requires a relatively large amount of glucose so that despite its relatively large water content (32%), water requirements of H-winter oil seed rape are in the same order as of wheat. The harvest index is another crop characteristic that has a high impact on ’transpirational’ water use. If crops show large investment in crop biomass, but have low yields, water requirements are large. Due to their large harvest indices, H-spinach, H-potato, H-maize (temperate climate), and H-grass show relatively low

Chapter 5 __________________________________________________________________________________________

54

water requirements, while H-wheat shows an intermediate requirement. In general, relatively small harvest indices, small water contents or large fat contents correspond with a relatively large water requirement, of which the requirement of winter oil seed rape forms a good example. Table 5.2 ‘Transpirational’ water requirements for the seven hypothetical crops (H-crops)

Crop ‘Transpirational’ water requirement Liters per kg

product Liters per kg dry

mass Liters per kJ

nutritional energy H-potato 38 150 0.011 H-wheat 204 240 0.014 H-maize (temperate climate) 145 171 0.010 H-maize (tropical climate) 120 142 0.008 H-field pea 121 185 0.011 H-winter oil seed rape 224 303 0.011 H-grass 14 176 0.010 H-spinach 11 140 0.022

55..55 ►► DDiissccuussssiioonn 5.5.1 Applicability Agricultural scientists have studied water needs of crops for specific situations and have provided useful tools for farmers for irrigation planning and water management. Those studies, however, do not separate water demand directly related to plant growth from site-specific losses. Temporal and spatial variation and the use of different system boundaries explain the large variation of water requirements found in literature. This thesis quantified one of the six relevant flows of the hydrological system, the ’transpirational’ water flow, under optimal conditions, with sufficient nutrients. The present thesis, therefore, gives low values compared to results of other studies that include the site-specific flows. The results should be considered as the absolute minimum water requirements to provide a yield. For practical situations, site-specific flows, such as soil evaporation or drainage, must always be taken into account. Although the results cannot be used to assess local water requirements in agriculture, they improve the insight into hydrological systems, demonstrate differences and similarities among crops, and provide a tool to evaluate the effect of an increase of global agricultural production on existing water resources. The calculations were done for hypothetical crops but, provided necessary data are available ’transpirational’ water requirements can be calculated for any crop. It is stressed that the data derived in this chapter are based on rough estimates of the environmental consequences of changes in the food production system and should not be interpreted at face value but as tools to better understand the system. The data show the direction of the changes and give an indication of their magnitudes. Although the chapter used many rough estimates, the analysis provides new insights into the functioning of the system. 5.5.2 Options to reduce water requirements by changing food consumption patterns If the fresh weight of crops is considered, the chapter showed large differences among ’transpirational’ water requirements. However, if nutritional energy is considered, requirements are all in the same order of magnitude, except for vegetables, represented here by H-spinach. Crops having small dry matter content, such as vegetables, make relatively large glucose investments in plant structural matter without nutritional energetic value for humans. In the calculations for water requirements, fibers were included, but in the calculations of nutritional energy they are excluded. The relatively large fiber content of vegetables compared to their macronutrient content explains the large water requirements per unit of nutritional energy. For two reasons, differences in water requirements among crops hardly form a basis for reduction options. Firstly, differences per unit of nutritional energy are low, so that there are only small differences among foods with the same function in human nutrition. Secondly, food consumption patterns must show variation to comply to nutritional constraints. A healthy diet provides nutrients in a balanced way, including vegetables with relatively large water requirements per unit of nutritional energy.


___________________________________________________________________________________________________________________________________________

Table 5.1 Main characteristics for seven hypothetical crops (H-crops) for ten categories: (i) low harvest index; (ii) high harvest index; (iii) low water content; (iv) high water content; (v) relatively low content of carbohydrates; (vi) relatively high content of carbohydrates; (vii) relatively low content of proteins; (viii) relatively high content of proteins; (ix) relatively low content of fats; (x) relatively high content of fats. The characteristics were derived from existing crops Crop Category Harvest index % Water

economic yield % Water rest fraction

Carbohydrates product (g kg-1)

Fats product (g kg-1)

Proteins product (g kg-1)

Nutritional energy (kJ kg-1)

H-potato ii, iv, vi, vii, ix

0.70 a. 75.0 a. 87 h. 190 g. 0 20 3570

H-wheat

i, iii, vi, ix 0.42 a. 15.0 a. 15 h. 693 f. 2 f. 117 f. 14364

H-maize

ii, iii, vi 0.45 a. 15.0 a 87 h. 710 f. 38 f. 92 f. 14450

H-field pea

i, viii, ix 0.51 b. 34.5 b. 87 h. 430 g. 15 g. 210 g. 11319

H-winter oil seed rape

i, x 0.32 c. 26.0 c. 87 h. 100 c. 420 c. 220 c. 21252

H-grass

ii, v, vii, ix 1.00 d. 92.0 d. - 54 d. 0 d. 30 d. 1411 d.

H-spinach

ii, v, vii, ix 0.90 h. 92.0 e. 92 e. 40 e. 0 e. 20 e. 510 e.

a. Goudriaan et al. (2001) b. Lecoeur and Sinclair (2001) c. Habekotté (1997) d. Centraal Veevoederbureau (1997) e. Voedingscentrum (1998) f. Catsberg and Kempen-van Dommelen (1997) g. Voorlichtingsbureau voor de Voeding (1973) h. Estimated value

Chapter 5 ___________________________________________________________________________________________________________________________________________

Fig. 5.3. Flow chart and necessary inputs for the calculation of ‘transpirational’ water requirements for crops. Agricultural data are combined with meteorological data. The output of one step is the input of the next one. This results in the value for ‘transpirational water’ requirements (liters kg-1) for a specific crop.

INPUTS INPUTS STEPS

• chemical composition harvest and rest fraction

• conversion factors for synthesis of plant materials from glucose (g g -1)

Step 6 total amount of ’transpirational’ water for food (liters kg-1)

• dry matter per MJ

global radiation • conversion

factor for carbohydrates (g per g glucose)

Step 5

total amount of glucose required to produce food

global radiation (MJ m-2)

Step 1 amount of glucose available for growth (g MJ-1 global radiation)

Step 2 amount of water for growth (liters MJ-1 global radiation)

Step 3 amount of water for glucose available for growth (liters g-1)

Step 4 total amount of biomass required to produce food

harvest index


__________________________________________________________________________________________

57

5.5.3 Increasing global food production The expected growth of the world population and the increase of per capita food demand in the coming decades might require a doubling of global food production. Under rainfed conditions, about 11 liters of ‘transpirational’ water is needed for the production of a MJ of nutritional energy. A doubling of global food demand implies in the first place a doubling of the critical requirements or ‘transpirational’ water. If the doubling is achieved via the route of the doubling of agricultural land areas, not only ’transpirational’ water but also site-specific losses increase, implying a doubling of total water requirements. The agricultural land area is limited, however. Given the potential production capacities (Penning de Vries et al., 1995), a doubling of yield levels is more likely. This means that globally only ‘transpirational’ water requirements, a minor part of total losses, double. On a global scale level, average evapotranspiration is 470 mm per year (Goudriaan et al., 2001). The average global wheat yield in 2002 was 2.7 tons per hectare (FAO, 2003b). Based on the results of this chapter presented in Table 5.2, this global yield requires 55 mm of ’transpirational’ water or 12% of the average, annual evapotranspiration. A doubling of wheat yields requires 110 mm of ‘transpiratonal’ water, or 23% of the average, annual evapotranspiration. The example shows that the demand for water expressed per unit of land area increases with increasing yield levels but also that ‘transpiratonal’ water flows are a minor part of total, annual evaporation losses. The possibility to increase yields depends on local factors that determine the availability of sufficient water. In some regions, this surplus will not be available, so that for the satisfaction of demand, yield levels in other regions must rise more than twice. In the light of the predicted future water scarcity, the replacement of foods of vegetable origin by foods with lower water requirements is no option to reduce the overall use of water. Better options are the increase of yield levels and harvest indices, or the reduction of waste flows and losses in the life cycle. This requires further research from various disciplines, especially the agricultural and hydrological sciences. 55..66 ►► CCoonncclluussiioonnss Site specific water flows are highly variable in time and among locations and form the major part of the hydrological system. The growth related water flow is a minor part of the system but quantifiable from a crop yield perspective. The linearity between this flow and attained yields provides a new insight into the food production system. Although ’transpirational’ water requirements per unit of fresh weight differ considerably among crops, requirements per unit of nutritional energy show minor variation. The main reason is that all green plants use the same basic strategy, photosynthesis, to produce biomass transpiring water at the same time. Differences in the efficiency to use glucose for the economic yield, therefore, is the main reason for variation in ‘transpirational’ water requirements among crops. This efficiency is mainly determined by the harvest index and the chemical composition of the dry matter. Crops rich in proteins have relatively large water requirements per unit of dry mass. Crops rich in oil show large water requirements, but this is compensated by their large nutritional energy content. Qualitative changes in food consumption patterns, therefore, do not result in substantial changes in annual, per capita requirements for water. By distinguishing between site-specific and crop growth specific water flows, this chapter quantified a central flow of the hydrological system, the water flow that actually passes a crop and that is directly related to the photosynthesis process. If yields increase, this water flow increases with the same factor. The results differ from results of traditional agricultural studies because they show the critical water requirements. These requirements are the only quantifiable, constant flows of a hydrological system. However, they must always be used in combination with locally, variable water needs. The results have two important consequences for food policy issues. Firstly, the chapter shows only small differences in water requirements among crops. Secondly, results indicate that under rainfed conditions, a doubling of food production on existing land areas does not imply a doubling of water use but only of ‘transpirational’ water use. This flow forms a minor part of total water requirements in agriculture. The assessment of ’transpirational’ crop water requirements provides insight into essential characteristics of the hydrological system and makes an important contribution to the information need relevant for global food security.

58

59

CChhaapptteerr 66

FFoooodd ccoonnssuummppttiioonn aanndd eeccoonnoommiicc ddeevveellooppmmeenntt,, aa ssppaattiiaall aanndd

tteemmppoorraall ccoommppaarriissoonn∗∗

Abstract The coming decades will bring enormous challenges concerning food security. The current growth of

the world population requires the production of more food. Along with population growth, many countries have

shown increased purchasing power causing a demand for more and other food. Studies on human nutrition have

shown that throughout the world a nutrition transition is taking place in which people shift towards more affluent

food consumption patterns. Especially in Asia, average per capita food supply increases, accompanied by shifts

towards the consumption of more meat and oil, and a decrease of the consumption of staples. The objective of

this chapter was to analyze the relationship between per capita income (GDP), and food supply, the composition

of food consumption, as well as the contribution of animal foods. It made a spatial comparison of food

consumption patterns of countries in various stages of economic development for the year 2001. Additionally, the

chapter made two temporal comparisons, a three century time trend for western Europe, and a forty year trend

for southern Europe. The spatial and temporal comparisons showed the same patterns of change. For low

incomes, an increase of GDP per capita paralleled changes towards the food consumption patterns of western

countries, characterized by a large gap between supply and actual consumption. Total supply (kilocalories per

capita per day) differed by a factor of two between low and high GDP. A second characteristic of changes of

consumption was an exchange of the fraction of nutritional energy from carbohydrates to fats and to animal

foods, while the protein fraction was stable. People with low GDP derived nutritional energy mainly from

carbohydrates. The contribution of fats to nutritional energy was small, of protein the same as for high GDP, and

that of animal sources negligible. People with high GDP derived nutritional energy mainly from carbohydrates and

fats, the contribution of animal sources was substantial. Whenever and wherever GDP increased, food supply,

the composition of consumption, and the contribution of animal foods all moved in the same direction. The fastest

changes occurred in the lowest income categories, below 5000 1990 International Geary-Khamis dollars. The

transition is completed at an annual, per capita income level of about 12 500 dollars. These findings have

important consequences for food security. The European transition occurred in a gradual way enabling

agriculture and trade to keep pace with the growth of demand. A continuation of present economic trends might

cause a large pressure on the food system within ten years because changes in food demand occur much faster

than projections indicate. Especially economic development in Asia will cause additional pressure on the global

food system.

66..11 ►► IInnttrroodduuccttiioonn

Today, the need to supply sufficient food for the growing world population is an important issue (Pinstrup-Andersen, 2000). One of the principle aims of the United Nations Millennium Development

∗ This chapter is a slightly adapted version of Gerbens-Leenes, P.W., Nonhebel, S. Food consumption and economic development, a spatial and temporal comparison. Submitted to Food Policy, June 2006.

Chapter 6 __________________________________________________________________________________________

60

goals, for example, is to eradicate hunger in the period 1990-2015 (United Nations, 2005). However, there is a difference between the supply of sufficient food and the consumption of food to prevent hunger. Supply is defined here as the average, annual, per capita availability of food commodities in a country. Many food commodities are basic ingredients for foods and have to pass chains and webs of a production system before they are available for consumption. In this system, several processes take place, such as, for example, processing sugar beet to manufacture sugar, or baking cakes (Gerbens-Leenes et al., 2003). The output of a food production system is the large quantity of food items available for consumption. Food consumption is defined here on two levels of scale, the household level and the per capita level. Household consumption concerns the foods available for a household, per capita consumption concerns the food that is actually eaten. In general, there is a gap between supply and actual consumption because in food production systems, as well as in households, losses occur. Food consumption shows repeated arrangements referred to as food consumption patterns (Gerbens-Leenes and Nonhebel, 2002). These patterns depend, among other things, on household income. When income increases, people spend more money on food (Pindyck and Rubinfeld, 2005; Vringer and Blok, 1995) and in this way change their food consumption. Income expressed as average, per capita Gross Domestic Product (GDP) differs among countries and in time (Maddison, 2003). In the period 1700-2000, for example, western Europe showed a twenty fold increase of GDP per capita; in the period 1961-2001, southern Europe observed a three to four fold increase. At the same time, food supply increased substantially (Fogel and Helmchen, 2002; FAO, 2004a) and food consumption patterns shifted towards patterns including more affluent foods such as meat, fats, beverages, and fruits, in combination with a declining consumption of staples (Whitney and Rolfes, 1999; Smil, 2002; Gerbens-Leenes and Nonhebel, 2002; Jobse-van Putten, 1995; Receveur et al., 1997; Mennell et al., 1992; Penning de Vries et al., 1995). These shifts cause changes in the composition of food consumed and related food supply. Today, the main differences among GDP per capita, food supply, and food consumption occur between developing and developed countries. GDP per capita in developing countries is low (Maddison, 2003) and food supply and consumption are mainly based on staples (FAO, 2003a). Developed countries show high GDP per capita, large supply and affluent food consumption patterns (Gerbens-Leenes and Nonhebel, 2002). However, studies on food consumption patterns in developing countries have reported shifts towards the affluent patterns of the western world (Grigg, 1995; Popkin, 2002; Lang, 2002; FAO, 2003a). The production of food requires natural resources such as agricultural land, water, and energy. More affluent food consumption patterns have larger requirements for these natural resources than patterns based on staples. Chapter 3, for example, has shown that a consumption pattern based on staples requires six times less land than an affluent pattern that includes large amounts of fats, beverages, and foods from animal sources, such as milk, cheese, and meat. In the next ten years, some developing countries will show large economic growth (World Bank, 2005b) and thus an increase of GDP per capita. Since the pressure on resources is already large, there is a need to quantify shifts in per capita food supply, and changes of consumption towards more affluent food consumption patterns that go along with an increase of GDP per capita. The specific aims of this chapter were twofold. First, to identify and analyze trends in per capita food supply and consumption that go along with increasing income (GDP per capita). The chapter quantified the gap between average, daily per capita food supply and actual consumption, shifts in the composition of food consumption, and shifts in the contribution of animal foods to consumption. Second, it identified regions where large changes in food supply and consumption will occur in the next ten years. The paper is organized as follows. Firstly, Section l assessed spatial differences among food supply and consumption. Uncertainty and inaccuracy, however, go along with attempts to make an accurate analysis of a complex system, such as the food system. To assess the sensitivity of the spatial comparison and to confirm trends, the chapter made four additional analyses. Section ll evaluated the probability of missing substantial food sources. To confirm trends found in Section l, Section lll made a temporal comparison for western Europe. 66..22 ►► FFoooodd ssyysstteemmss Food systems include a production and a consumption system. Figure 6.1 shows a simplified overview of the system. A production system consists of (i) agricultural production made up of primary and secondary production that provide a supply of crop and animal commodities, (ii) the food industry that applies the supply from primary and secondary production to manufacture food items for

Food consumption and economic development, a spatial and temporal comparison

__________________________________________________________________________________________

61

consumption, and (iii) retailing. A consumption system consists of (i) household consumption defined here as types and amounts of food bought or produced by a household and (ii) per capita consumption that concerns the food that is actually eaten. Food items originate from production systems that often consist of several production chains made up of chain links. A production chain processes commodities from agriculture, the first chain link, into ingredients for food, and finally food items available for consumption (Gerbens-Leenes et al., 2003).

Fig. 6.1. Simplified overview of the food system that includes a production and a consumption system. A production system consists of agricultural production made up of primary and secondary production that provide a supply of crop and animal commodities, the food industry that applies the supply from primary and secondary production to manufacture food items for consumption, and retailing. A consumption system consists of household consumption defined here as types and amounts of food bought or produced by a household and per capita consumption that concerns the food that is actually eaten. Physical streams flow from agriculture to the food industry, retailing, households and finally to per capita consumption. Sometimes, streams flow in the opposite direction. These opposite streams concern waste streams that are reused in an earlier stage of production. In every chain link of the system, losses take place. 6.2.1 Agricultural production Globally, agricultural production provides only a limited amount of commodities important for the food system. The fifteen main categories of crop commodities are sugar cane, root crops, vegetables, maize, paddy rice, wheat, fruits, potato, sugar beet, cassave, soybean, barley, pulses, oil seed rape, and sorghum; the six main animal commodities are raw milk, pork, poultry, beef, mutton, and goat meat (FAO, 2004a). For almost all countries starting in 1961, the FAO food balance sheets (FAO, 2004a) provide information on annual, market supply of these commodities. Four components, water, and the macronutrients carbohydrates, proteins, and fats dominate the composition of commodities (FAO, 2004a; Whitney and Rolfes, 1999; Voedingscentrum, 1998a). The macronutrient content of commodities, such as wheat, soybean, or pork, is genetically determined so that all crop and animal commodities show a specific composition (kg macronutrient per kg dry matter) (Penning de Vries et al., 1989; Schmidt-Nielsen, 1988). Based on composition, the main commodities form four categories: (i) staples, crops that mainly provide carbohydrates and some proteins; (ii) protein rich crops, crops that provide proteins as well as carbohydrates; (iii) oil crops, crops that provide vegetal fats for the production of oil, and carbohydrates and proteins for feed (Penning de Vries et al., 1989); and (iv) animal commodities, commodities that provide high quality proteins and fats (Whitney and Rolfes, 1999). The composition of a commodity determines its suitability for food

Per capita consumption

Household consumption

Primary production

Retailing

Food industry

Secondary production

Chapter 6 __________________________________________________________________________________________

62

items and the function for human nutrition (Whitney and Rolfes, 1999). For example, people consume staple crops, such as wheat and potatoes, for their carbohydrates, pulses for proteins and carbohydrates, while oil crops, such as oil seed rape, provide vegetal oil. 6.2.2 Food industry The food industry selects and processes commodities from agriculture to manufacture food items (Catsberg and Kempen-van Dommelen, 1997). The food industry often separates commodities into several fractions mainly based on different composition characteristics, such as the content of fat, protein, and carbohydrate. Soybeans, for example, are split into an oil and an oil cake fraction (Kramer and Moll, 1995). Oil is a basic ingredient for margarines, oil cakes for livestock feed. The fractions form the basic ingredients for further production when industry joins and processes ingredients into food items. In the western world, technological developments in agriculture, transportation, and food conservation at the end of the 19th century stimulated the expansion of the food industry and food preparation shifted from households to industry (Jobse-van Putten, 1995). 6.2.3 Household consumption Household consumption concerns the foods available for a household, either bought in retailing, sometimes produced in homegardens (Fernandes and Nair, 1986; Pallot and Nefedova, 2003). In a household, food preparation into dishes and meals takes place. The repeated arrangements of consumption characterized by types and quantities of food items and their combination into dishes and meals are termed food consumption patterns (Gerbens-Leenes and Nonhebel, 2002). They depend on factors like preference, habit, seasonal availability of foods, household income, convenience, social relations, ethnic heritage, religion, tradition, culture, time constraints, and nutritional constraints. 6.2.4 Per capita consumption Per capita consumption or human nutrition concerns the food that is actually eaten, either from household food preparation or directly from retailing. Food surveys provide detailed information on nutrition (see also Appendix 2). For nutrition, the composition of food in terms of macronutrients is important because they provide energy and are essential for body functions. Humans can derive energy from different combinations of macronutrients, however. This flexibility contributes to variation among macronutrient composition of consumption and to variation among food consumption patterns. 6.2.5 Physical streams in the food system Figure 6.1 shows that in the food system, physical streams flow from agriculture to the food industry, retailing, households and finally to per capita consumption. Sometimes, streams flow in the opposite direction. These opposite streams concern waste streams that are reused in an earlier stage of production, for example manure from secondary systems in agriculture, or waste from the food industry for livestock feed (Nonhebel, 2004). In every chain link of the system, losses take place. For example, the food industry processes 1.4 kilograms of wheat to manufacture 1 kilogram of flour (Kramer and Moll, 1995). A study on household consumption in Sweden has estimated that during meals, 10 percent of the food remains behind on the plate and is wasted (Karlsson, 2001). These losses cause a gap between supply from agriculture and actual, per capita consumption. 66..33 ►► SSeeccttiioonn ll,, ssppaattiiaall ddiiffffeerreenncceess aammoonngg ffoooodd ssuuppppllyy aanndd ccoonnssuummppttiioonn 6.3.1 Introduction Section l Per capita food consumption depends, among other things, on income. When income increases, people spend more money on food (Pindyck and Rubinfeld, 2005; Vringer and Blok, 1995). To analyze the relationship among food supply, per capita food consumption, the contribution of animal foods to consumption, and income, this Section assessed spatial differences in 2001 for fifty two countries in different stages of economic development.


__________________________________________________________________________________________

63

6.3.2 Materials and methods Section l 6.3.2.1 Units of calculation Several studies have shown that natural resource use among food items and consumption arranged in food consumption patterns shows large differences (e.g. Van Engelenburg et al., 1994; Kok et al., 1993; Kramer and Moll, 1995; Gerbens-Leenes et al., 2002). Especially meat, fats, and beverages have relatively large requirements for energy and land (MJ kg-1, m2 kg-1). This means that the composition of food consumption has a large impact on natural resource use. For the comparison of trends of per capita food consumption related to average, per capita income, this chapter simplified consumption in terms of macronutrient composition (fats, carbohydrates, and proteins). For most countries, FAO food balance sheets provide information on per capita supply of nutritional energy (kilocalories per day from vegetal and animal products), and supply of proteins and fats (grams per day). The chapter expressed per capita food consumption in the fraction of nutritional energy provided by the macronutrients fats, carbohydrates, and proteins, the macronutrient energy percentage (E%). It expressed the contribution of food items derived from animal systems in the fraction of nutritional energy derived from animal sources (A%), and supply from agriculture in nutritional energy (kilocalories per capita per day). The E% and A% were calculated by

protein E% = %100*.*E

pkcalP (1)

fat E% = 100*.*E

fkcalF (2)

carbohydrate E% = %100*)).*().*((

EfkcalFpkcalPE +−

(3)

A% = %100*EA

(4)

where P is the average, daily supply of protein (grams); kcal.p the nutritional energy supply of protein (4 kilocalories per gram); E the average, daily per capita supply of nutritional energy (kilocalories); F the average, daily supply of fat (grams); kcal.f the nutritional energy supply of fat (9 kilocalories per gram); and A the average, daily per capita supply of nutritional energy from animal sources (kilocalories). The chapter derived data on per capita supply from FAO food balance sheets (FAO, 2004a) and values on nutritional energy for protein and fat from the Dutch Nutrition Council (Voedingscentrum, 1998a). The chapter assessed spatial variation among food supply, the contribution of animal foods, and the composition of consumption for fifty two countries in 2001. Appendix 1 shows an overview of these countries. The chapter selected countries from Africa, Asia, Eastern Europe, Latin America, the Middle East, and the OECD, in different stages of development, with more than five million inhabitants. These countries formed two clusters of developed and developing countries, however. It therefore added three smaller transition countries, clustered into a small country group. 6.3.2.2 Income, food supply and per capita consumption Per capita income depends, among other things, on the development status of an economy, the size of households, and on income distribution in a country. The World Bank (2005a) expresses economies in Gross Domestic Product (GDP). Information on average, GDP per capita is available for most countries from various sources. Maddison (2003), for example, has made a database of the historical, economic development of the world. From the Middle Ages onwards, that database provides information on the economic development status of almost all countries in the world on a national and per capita basis. The database expresses GDP in 1990 International Geary-Khamis dollars (G-K dollars). The Geary-Khamis method is an aggregation method in which international prices and a countries Purchasing Power Parity, depicting relative country price levels, are estimated simultaneously from a system of linear equations and expressed in G-K dollars (United Nations Statistics Division, 2006). This chapter applied average GDP per capita (1990 G-K dollars) as an

Chapter 6 __________________________________________________________________________________________

64

indicator for income and combined this with results on food supply and consumption. It derived data from Maddison (2003) since it covers both historic as well as global information. 6.3.3 Results and discussion Section l The spatial analysis generated three types of results. It showed the relationship between average, per capita income (GDP) on the one hand, and average, daily, per capita food supply, the composition of food consumption, and the contribution of animal foods to consumption, on the other. 6.3.3.1 Per capita income and food supply Per capita food supply depended on a countries’ annual, per capita income (GDP). Figure 6.2 shows that supply varied between 1600 kilocalories per capita per day for low GDPs and 3800 kilocalories for high GDPs, a difference of a factor of almost two and a half. The figure shows that especially for low GDPs, differences per unit of GDP were large, while for high GDPs differences were negligible and saturation occurred. For developed countries, Japan was an exception since it combined a high GDP per capita with relatively small supply.

0

1000

2000

3000

4000

0 5000 10000 15000 20000 25000 30000

Annual per capita GDP (1990 International Geary-Khamis dollars)

Nut

ritio

nal e

nerg

y su

pply

200

1 (K

iloca

lorie

s pe

r cap

ita p

er d

ay)

Fig. 6.2 Relationship between annual GDP per capita (1990 International Geary-Khamis dollars) and nutritional energy supply (Kilocalories per capita per day). The relationship was based on data from 57 countries in different stages of development in 2001. For an overview of countries, see Appendix B. 6.3.3.2 Per capita income, composition of consumption and contribution of animal foods Figure 6.3a shows the relationship between the macronutrient composition of consumption and annual GDP per capita. It shows that the fraction of nutritional energy (E%) provided by proteins (the protein E%) was stable, 9-13 E%. The carbohydrate and fat E%, however, showed a relationship with GDP. In countries with low GDPs, people derived nutritional energy mainly from carbohydrates, and a small fraction from fats. In Bangladesh, for example, the country with the lowest GDP in the analysis, people derived 80 percent of nutritional energy from carbohydrates, and 11 percent from fats. For consumption in countries with high GDPs, carbohydrates were less important and more energy was provided by fats. The average consumer in the U.S., France, and Denmark, for example, derived 45 to 50 E% from carbohydrates, and 40 E% from fats. The figure shows that for countries with low GDPs, differences in composition were large; for countries with high GDPs, differences were negligible and saturation occurred at the GDP level of Greece. The average consumer in that country derived 51 percent of its energy from carbohydrates and 36 percent from fats.


__________________________________________________________________________________________

65

6.3.a

0

20

40

60

80

100

0 5000 10000 15000 20000 25000 30000


Sour

ce o

f nut

ritio

nal e

nerg

y 20

01 (E

%)

fatE%proteinE%carbohydrateE%

6.3.b

0

20

40

60

80

100

0 5000 10000 15000 20000 25000 30000


Nut

ritio

nal e

nerg

y fr

om a

nim

al s

ourc

es (A

%)

Fig. 6.3.a-b. Fig. 6.3a shows the relationship between annual GDP per capita and the composition of food consumption patterns in terms of the fraction of nutritional energy derived from fat (fat E%), from protein (protein E%), and from carbohydrate (carbohydrate E%). Fig. 6.3b. shows the relationship between annual GDP per capita and the composition of food consumption patterns in terms of the fraction of nutritional energy from animal sources (A%). The relationships were based on data from 57 countries in different stages of development in 2001. Figure 6.3b shows the relationship between the fraction of nutritional energy derived from animal sources (A%) and annual GDP per capita. For consumption in countries with low GDPs, A% is almost negligible; for consumption in countries with high GDPs, the fraction is about 25-40 percent.

Chapter 6 __________________________________________________________________________________________

66

Consumption in Bangladesh, for example, showed an A% of only 3 percent, while consumption in Denmark had an A% of 40 percent. The figure shows that for low GDPs, differences were large; for high GDPs, differences per unit GDP were smaller. It is stressed, however, that A% indicates the fraction of energy derived from meat, dairy, and eggs of food consumption and not amounts of foods consumed. Some countries with high GDPs, for example Canada and the U.S., showed consumption with relatively small fractions of energy derived from animal sources, 27 and 30 E%. For the average Canadian, the supply of meat was 101 kilograms per capita per year and of raw milk 204 kg. For the average U.S. citizen, the supply of meat was 121 kilograms and of raw milk 262 kg. In an OECD country with a lower GDP than the U.S. and Canada, the Netherlands, on the other hand, people derived 34 percent of nutritional energy from animal sources, but the supply of meat was much smaller than in the U.S. or Canada, only 90 kilograms per capita per year, while the consumption of raw milk was larger, 336 kilograms (FAO, 2004a). The example shows that the fraction of energy derived from animal sources, does not increase infinitely. Jobse-Van Putten (1995) has also shown that in the western world, high income groups consume less meat that low income groups. In general, animal protein has a better quality than vegetal protein (Whitney and Rolfes, 1999). For all countries, the protein E% of per capita consumption was about the same. An increase of the fraction of nutritional energy derived from animal foods, therefore, does not imply an increase of the fraction of protein but rather an improvement of protein quality. Especially for developing countries, this improves the quality of consumption. Figure 6.3a and b show that in some countries, consumers deviated from trends. In Japan, for example, a high GDP per capita paralleled a relatively small per capita food supply, while the composition of consumption resembled a pattern related to a lower GDP. This indicates that also other factors than GDP, for example, cultural factors, affect consumption. 6.3.3.3 Uncertainty and inaccuracy of results Three factors caused uncertainty and inaccuracy of results. These were: (i) data quality; (ii) the use of average data; and (iii) the use of supply data. Data quality The chapter derived data on food supply from FAO food balance sheets (FAO, 2004a), and on GDP from Maddison (2003). The FAO obtains data from national datasets. This means that data for different countries probably have different qualities, dependant on the degree of development of national, statistical organizations. Within countries, data quality can differ among years. Major events, like political instability, or improvements of methods of statistical organizations affect data quality. Even for countries with high quality statistical organizations, different sources provide different data. For the Netherlands in 2000, per capita butter consumption, for example, differed by a factor of three among datasets. According to the FAO, the Dutch consume 2.1 kg of butter per capita per year (FAO, 2004a); according to the Statistics Netherlands (CBS) 3.3 kg (LEI-DLO/CBS, 2002); and Eurostat estimates 6.8 kg (LEI-DLO/CBS, 2002). Maddison (2003) also derived data from national statistics. That dataset probably has the same drawbacks as the FAO dataset. The use of average data The second factor that contributed to uncertainty and inaccuracy was the use of average data. Per capita data derive from information on a national level, and are therefore average numbers. In some countries, income distributions are large and differences in food consumption occur among population groups. These differences are not reflected in national data. This means that the use of average data underestimates trends found in here. The use of supply data The third factor for uncertainty and inaccuracy for the comparison was the use of FAO supply rather than consumption data. Before food is available for per capita consumption, commodities and foods pass complete food chains, from farm to fork, in which a number of processes takes place to produce the final foods. In all chain links and transportation between links, losses take place. The size of the gap between supply and per capita consumption is not precisely known, but it can be argued that long chains have larger losses than short chains. Sometimes, people produce food outside the market, for example in home gardens (Fernandes and Nair, 1986; Pallot and Nefedova, 2003). The FAO food balance sheets provide supply data on a national level and do not take nonmarket production into account (FAO, 2004a). For low GDPs, the chapter found a ratio of supply over physical requirements of about 1.0. For high GDPs, however, the ratio was larger, about 1.8, indicating that about half of


__________________________________________________________________________________________

67

supply is not eaten at all. By excluding nonmarket production, the chapter underestimated supply. Neglecting losses in food chains could have had effects on the assessments. 6.3.4 Conclusions Section l Differences in annual GDP per capita among countries paralleled differences in the composition of per capita food consumption. In all countries investigated, the fraction of nutritional energy derived from protein was 9-13 E%. For consumption in countries with low GDPs, the fraction from fat was about 11 E%, from carbohydrate 80 E%, and the fraction from animal sources 3 percent. For consumption in countries with high GDPs, the fraction from fat was 35-40 E%, from carbohydrate 45-50 E%, and the fraction from animal sources 30-40 percent. Total supply ranged between 1600 kilocalories per capita per day for countries with low GDPs, and 3800 kilocalories for countries with high GDPs. The largest differences per unit of GDP occurred for relatively low GDPs, i.e. below 5000 1990 International Geary-Khamis dollars (G-K dollars) per capita per year. GDP levels between 5000 and 12 500 G-K dollars showed smaller differences. Above a GDP level of 12 500 G-K dollars, supply did not increase and the composition of consumption and contribution of animal foods to consumption did not show large differences among countries. 66..44 ►► SSeeccttiioonn llll,, tthhee uussee ooff ssuuppppllyy ddaattaa 6.4.1 Introduction Section ll For the comparison of the composition of per capita consumption and the contribution of animal foods, Section l derived supply data from the FAO food balance sheets. To analyze the use of supply rather than per capita consumption data, Section ll compared differences in composition between supply and per capita consumption. Next, it evaluated the size of the gap between supply and actual, per capita consumption in relation to per capita income (GDP). 6.4.2 Materials and methods Section ll Information on market supply of commodities is available for most countries. Information on per capita consumption, i.e. the amount and types of food that people actually eat, is scarcer, however. Food surveys provide information on nutritional characteristics of per capita consumption for a representative population group in a specific country (FAO, 2005). They provide data on nutritional energy intake (kilocalories per capita per day), and most surveys provide data on of the fraction of nutritional energy from the macronutrients fats, carbohydrates, and proteins, the macronutrient energy percentage (E%). Some surveys make a distinction between rural and urban population groups. A number of surveys has been performed in developing countries (FAO, 2005), two time series are available for developed countries, the Netherlands (Voedingscentrum/TNO, 1998b) and the United States (USDA, 2005). Appendix B shows an overview of these surveys. To analyze differences between FAO supply data used in Section l and actual, per capita consumption data, this section compared the composition of per capita consumption and related supply. Every commodity has a specific composition in terms of macronutrients, such as fat E%. Results of Section l showed that the contribution of protein to nutritional energy of consumption, the protein E%, was stable but that differences occurred between fat and carbohydrate E%, reflecting a difference in the composition of consumption. This Section assumed that a difference in fat E% between supply and per capita consumption reflects a difference of composition. For the comparison, the section used information on the fat E%. It obtained data on fat E% from eighteen food surveys. Appendix 2 marked these surveys with an asterix *. The section calculated the fat E% of related supply using equation 2 from Section l and derived data from the FAO (2004a). Section l showed that food supply increases along with increasing GDP. To evaluate the size of the gap between supply and actual, per capita consumption, Section ll assessed the relationship between nutritional energy intake (kilocalories per capita per day) and average, annual GDP per capita. It combined data from thirty one food surveys from twenty six countries (see Appendix B) with information on GDP from Maddison (2003).

Chapter 6 __________________________________________________________________________________________

68

6.4.3 Results and discussion Section ll 6.4.3.1 The composition of supply and consumption Figure 6.4 shows that the fat E% of supply and actual, per capita consumption were in the same order of magnitude, indicating that macronutrient compositions were the same. Eventual losses or nonmarket production did not cause shifts in composition. This justified the use of supply rather than consumption data for the analysis of per capita consumption in Section l.

Fig.6.4. Comparison between the fraction of nutritional energy derived from fat (fat E%) of actual consumption and of related supply. Data on fat E% of consumption derived from food surveys, the fat E% of related supply was calculated from FAO food balance sheets. 6.4.3.2 Per capita income and nutritional energy intake Available food surveys were performed in countries with large differences in per capita income (GDP). Sixteen food surveys indicated nutritional energy intakes between 2000 and 2500 kilocalories per capita per day, which is in the range of actual requirements (Whitney and Rolfes, 1999). Eight surveys indicated intakes between 1700 and 2000 kilocalories, while five studies report intakes over 2500 kilocalories per capita per day. This thesis found no relationship between nutritional energy intake and annual GDP per capita, however. For the Netherlands and the U.S., two countries with high GDPs, energy intakes were between 2000 and 2500 kilocalories per capita per day. This is in the same order of magnitude than intakes in countries with low GDPs. Twelve food surveys have made a distinction between rural and urban consumption but did not find substantial differences among energy intakes. It can be argued that the physical requirement of the human body determines the quantity of nutritional energy intake. Section l showed that the increase of GDP per capita went along with larger supply. The comparison of food surveys showed that per capita consumption did not increase. The increase of the gap between supply and per capita consumption, therefore, is likely to result from larger supply. 6.4.3.3 Uncertainty and inaccuracy of results Section l discussed uncertainty and inaccuracy caused by the use of the FAO and Maddison datasets. Another reason for uncertainty and inaccuracy is that the chapter derived information from food surveys that were probably all performed in a different way, generating different types of inaccuracies and uncertainties. For example, people tend to underreport consumption (CBS, 1994; Kok et al., 1993). The way food surveys have addressed this effect has probably differed among

0

10

20

30

40

0 10 20 30 40

Fat E% food surveys

Fat E

% fo

od b

alan

ce s

heet

s


__________________________________________________________________________________________

69

surveys, generating different inaccuracies. The lowest value was for the Philippines, 1800 kilocalories per capita per day, which is below physical requirements, the highest for Jordan, 3200 kilocalories per capita per day, far above physical requirements. The examples show that methods among surveys have probably differed generating under and overestimations. 6.4.4 Conclusions Section ll It can be concluded that for the assessment of changes in the composition of per capita food consumption and the contribution of animal foods to consumption, the use of FAO supply data did not lead to other results than the use of actual consumption data. GDP increase did not have an effect on nutritional energy intake, which is stable and reflects physical requirements. The increase of food supply (kilocalories per capita per day), therefore, is not caused by an increase of consumption but reflects an increasing gap between supply and per capita consumption along with increasing income. 66..55 ►► SSeeccttiioonn llllll,, tteemmppoorraall ddiiffffeerreenncceess aammoonngg ffoooodd ssuuppppllyy aanndd ccoonnssuummppttiioonn 6.5.1 Introduction Section lll Over the last millennium, Europe has shown continuous economic growth (Maddison, 2003). Between 1700 and 2000, for example, in France, average GDP per capita increased from 900 to 21 000 1990 International Geary-Khamis dollars (G-K dollars), in Great Britain, from 1250 to 20 000 G-K dollars. Between 1961 and 2001, Italy, Greece, Spain, and Portugal showed a three to four fold increase of average GDP per capita. These periods went along with large changes of food consumption (Jobse-Van Putten, 1995; Fogel and Helmchen, 2002; FAO, 2004a). Section l analyzed the effect of income (GDP) on food supply and per capita consumption for countries in different stages of development in 2001. To evaluate if trends also exists within countries, Section lll made two time trends, for per capita supply in France and Great Britain, and for supply and consumption in southern Europe. 6.5.2 Materials and methods Section lll Studies of historic food consumption often describe changes in a qualitative way (e.g. Mennell et al., 1992; Jobse-Van Putten, 1995) and do not provide quantitative data. An exception is the analysis of Fogel and Helmchen (2002) that has quantified nutritional energy supply for France and Great Britain between 1700 and 2000 (kilocalories per capita per day). To evaluate per capita food supply over time, the section first made a three century time trend for supply in France and Great Britain. It combined data (kilocalories per capita per day) from Fogel and Helmchen (2002) with GDP data (G-K dollars) from Maddison (2003). Second, the section made a four decade time trend (1961-2001) for food consumption in southern Europe. For Italy, Spain, Portugal, and Greece, it assessed the increase of per capita supply, changes in the composition of food consumption, and changes in the contribution of animal foods. For the comparison of composition, it applied equations 1-3 from Section l, for the comparison of the contribution of animal foods, it applied equation 4. The section derived data from the FAO (2004a). Next, it combined data on supply (kilocalories per capita per day), results on composition (macronutrient E%), and contribution of animal foods (A%) with data on income (GDP in G-K dollars) from Maddison (2003). 6.5.3 Results and discussion Section lll 6.5.3.1 Food supply in France and Great Britain, 1700-2000 Figure 6.5 shows that in France and Great Britain, increasing GDP per capita paralleled larger food supply. Per capita supply doubled over the three centuries considered, from 1700 kilocalories per capita per day in 1700 to 3500 kilocalories in 2000. The largest increase per unit of GDP occurred for values below 5000 G-K dollars, above this level, the increase gradually slowed down. The figure also shows the function based on the spatial analysis of Section l. Results for food supply in France and Great Britain were in the same order of magnitude than results of the spatial analysis.

Chapter 6 __________________________________________________________________________________________

70

Development of nutritional energy supply in France and Great Britain 1700-2000

0

500

1000

1500

2000

2500

3000

3500

0 5000 10000 15000 20000


Nut

ritio

nal e

nerg

y su

pply

(kilo

calo

ries

per c

apita

per

day

)

historical data

function based on 2001data

Fig. 6.5. Relationship between annual GDP per capita (1990 International Geary-Khamis dollars) and nutritional energy supply (Kilocalories per capita per day) for France and Great Britain between 1700 and 2000. The figure also shows the function based on the regional analysis presented in Figure 6.2.

6.5.3.2 Food supply and consumption in southern Europe, 1961-2001 Figures 6.6a-c show results for southern Europe in the period 1961-2001. Figure 6.6a shows that per capita, daily, nutritional energy supply increased from 2500 kilocalories per capita per day for a GDP of 3000 G-K dollars (Portugal, 1961) to 3700 for a GDP of 12 500 G-K dollars (Greece, 2001). Figure 6b shows changes of the composition of food consumption. For protein, the E% varied between 13 and 18 E%, but was independent of GDP. For carbohydrate and fat, changes occurred. Increasing GDP paralleled a decrease of the fraction of nutritional energy derived from carbohydrates, and an increase from fats. For GDPs below 5000 G-K dollars, the carbohydrate E% was 60-70 E%, the fat E% 20-30 E%. The composition stabilized at an annual GDP of 12 500 G-K dollars. At this level, people derived 45 E% from carbohydrates and 40 E% from fats. Figure 6c shows the increase of the fraction of energy from animal sources (A%). A% was relatively small for low GDPs, the smallest value was 13 A% for a GDP of 3400 G-K dollars (Greek consumption, 1961). For a GDP of 14 200 G-K dollars (Portuguese consumption, 2001), the A% was 30 percent. Although GDPs of southern Europe fell outside the trajectory below 3500 and above 19 000 G-K dollars, when figure 6a-c was compared to Figure 6.2a-c, results were in the same order of magnitude than results of the spatial analysis. 6.5.3.3 Comparison of results with information from food surveys To evaluate if information from food surveys could confirm trends found in here, the chapter compared the fat E% of per capita consumption derived from eleven surveys in developing countries that made a distinction between urban and rural patterns. Figure 6.7 shows the results. Except for Egyptian consumption, urban per capita consumption had a larger fat E% than rural consumption. The surveys were done in countries with relatively low GDP per capita, i.e. within the trajectory where large differences in composition of food consumption occur per unit of GDP. It is likely that GDPs were higher for urban populations, a factor that explained the difference of fat E%. This confirmed results of this chapter. 6.5.3.4 Uncertainty and inaccuracy of results All factors that contributed to uncertainty and inaccuracy of results in Section l and ll also contributed to uncertainty and inaccuracy of Section lll. Moreover, the historical analysis of Fogel and Helmchen (2002) has reported average, daily, nutritional energy intakes below physiological requirements. In


__________________________________________________________________________________________

71

general, nutritional energy requirements are constant per unit of body mass (Whitney and Rolfes, 1999). Average energy intakes below the physiological requirement, might be possible, however, when the fact that three centuries ago, people were smaller and had more children, with less body mass, is taken into account.

6.6a

0

1000

2000

3000

4000

0 5000 10000 15000 20000


Nut

ritio

nal e

nerg

y su

pply

(kilo

calo

ries

per c

apita

per

day

)

Total calories

trendline 2001

6.6b

Development macronutrient supply Southern Europe 1961-2001

0

20

40

60

80

100

0 5000 10000 15000 20000


Com

posi

tion

of fo

od c

onsu

mpt

ion

patte

rns

( mac

ronu

trie

nt

ener

gy%

)

fatE%proteinE%carbohydrateE%

Chapter 6 __________________________________________________________________________________________

72

6.6c

0

20

40

60

80

100

0 5000 10000 15000 20000

Annual per capita GDP 1990 International Geary-Khamis dollars)

Nut

ritio

nal e

nerg

y fr

om a

nim

al s

ourc

es (%

)E% animal sources

trendline 2001

Fig. 6.6a-c. Fig. 6a shows the relationship between annual GDP per capita (1990 International Geary-Khamis dollars and nutritional energy supply (kilocalories per capita per day) for southern Europe between 1961 and 2001. The figure also shows the function based on the spatial analysis of Section l. Fig. 6.6b. shows the relationship between annual GDP per capita and the composition of food consumption patterns in terms of the fraction of nutritional energy derived from fat (Fat E%), from protein (Protein E%), and from carbohydrate (Carbohydrate E%) for southern Europe between 1961 and 2001. Fig. 6.6c shows the relationship between annual GDP per capita and the composition of food consumption patterns in terms of the fraction of nutritional energy from animal sources (%) for southern Europe between 1961 and 2001. The figure also shows the function based on the spatial analysis of Section l. The relationships were based on data from Italy, Greece, Spain, and Portugal.

0

5

10

15

20

25

30

35

Mali 19

94

Zimbab

we 198

5

Banglad

esh 19

95

China 198

2

China 199

2

Panam

a 199

2

Trinidad

and Tobag

o 1992

Egypt 2

000

Iran 19

83

Iran 19

92

Jord

an 19

97

Fat E

% fo

od c

onsu

mpt

ion

patte

rn

urban rural

Fig. 6.7. Fraction of nutritional energy derived from fat (fat E%) for urban and rural populations in nine countries in different years. The comparison was based on data from eleven food surveys in developing countries (for an overview, see Appendix B).


__________________________________________________________________________________________

73

6.5.4 Conclusions Section lll France, Great Britain, and southern Europe showed the same trends as obtained in the spatial analysis. Higher GDPs favored larger supply, larger fractions of nutritional energy from fats, and larger contribution of animal foods; at the same time, the fraction derived from carbohydrates decreased. The fraction of energy from protein was stable. The largest changes occurred for relatively low GDPs, above an annual GDP per capita of 12 500 G-K dollars, food supply and per capita consumption were stable. 66..66 ►► GGeenneerraall ddiissccuussssiioonn 6.6.1 Trends The general impression that in western countries per capita food consumption expressed as nutritional energy intake has increased over the last decades is not in accordance with results presented here. Nutritional energy intake is far more constant than food supply or the composition of consumption, and reflects physical requirements. Along with increasing income, the size of the gap between actual, per capita consumption and supply (kilocalories per capita per day) increases, not physical consumption. An explanation is that affluent countries with high GDPs also have better developed food industries with longer food chains, and probably larger losses, than countries with low GDPs. Another explanation is that in countries with low GDPs, where food is more scarce, people prepare and consume food more efficiently and in this way generate less and smaller waste streams. The FAO food balance sheets do not provide information to confirm this hypothesis, however. The evaluation of the increasing gap between supply and per capita consumption that parallels increasing GDP requires further research. Results of the spatial and temporal analyses all show the same patterns of change. The largest changes in food supply and per capita consumption occurred for relatively low GDPs, below 5000 G-K dollars, between 5000 and 12 5000 dollars changes were relatively small, above an annual GDP per capita of 12 500 G-K dollars, food supply and the composition of consumption were stable. This is in accordance with many studies that have shown that increasing societal affluence causes shifts in the consumption of specific foods and commodities. By simplifying per capita consumption further than assessments of existing studies, and by adding the aspect of economic development stage, the thesis identified strong similar trends. However, it should be realized that this chapter addressed composition rather than absolute amounts of food. As a result, some countries with high GDP showed a relatively low contribution of animal foods while absolute consumption was large. Although uncertainties occurred, all analyses showed similar directions of change. Despite the use of rough estimates, differences among countries, developments in time, as well as differences between urban and rural populations, all showed the same trends. It is stressed, however, that results obtained here cannot be interpreted at face value. They give an indication of the direction of changes of food supply, the composition of consumption, the contribution of animal foods, and their magnitudes. In combination with estimates of GDP increases, the method presented here provides a tool to quantify these changes, and indicate where and when they will probably take place. 6.6.2 Future changes The most important finding of this paper is that the main changes occur for per capita, annual incomes below 12 500 G-K dollars. If trends found in here are also valid for the future, this has important consequences for food security in the coming decade. Today, about 85 percent of the total world population lives in six regions: (i) the OECD countries; (ii) Latin America; (iii) Africa; (iv) China; (v) India; and (vi) the rest of Asia. Table 1 shows the nutrition, GDP and population characteristics of these regions. In four regions, per capita income levels (GDP) are below 5000 dollars per year, i.e. within the trajectory where the largest changes in food supply, composition of consumption, and contribution of animal foods occurred. China, India, and the rest of Asia combine low GDP per capita with large growth rates. This means that in the next ten years, large changes are likely to occur in Asia. If the Asian countries maintain economic growth along existing lines, the period 2006-2015 might show a substantial increase of per capita food supply, while the composition of consumption might shift towards the affluent patterns of western countries, characterized by large consumption of fats and animal foods, and small consumption of staples. For Latin America and Africa, economic growth will

Chapter 6 __________________________________________________________________________________________

74

probably be small. Here, population growth will be the main driver for increasing total food demand. For the OECD, no substantial changes are likely because food consumption in these countries have already reached their saturation levels and population size is more or less stable. Table 6.1 Per capita nutrition characteristics in 2001, GDP characteristics (Geary-Khamis dollars), expected national GDP growth, and population characteristics for six regions that include 85% of the global population

Nutrition characteristics 2001b. GDP characteristicsc Population characteristics

Region

Energy supplyd..

Fat E%

Protein E%

Energy from animal sources (%)

Annual, per capita GDP 2001

National GDP growth

Annual, GDP per capita 2015

Size 2001 (billion)e

*

Annual growthf

Size 2015 (billion)**

China 2953 26 11 20 3800 6,9% 9671 1,29 0.7% 1.42

India 2385 19 9 8 1926 4,0% 3335 1,03 1.4% 1.25

OECD 3493 36 12 27 21538 1,8% 27648 0,89 0.4% 0.94

Asiaa 2540 18 9 9 2760 2,1% 3692 0,76 1.3% 1.20

Africa 2519 18 10 7 1615 0,6% 1756 0,52 2.6% 0.74 Latin America 2905 26 11 20 6174 1,2% 7296 0,45 1.3% 0.54

a. Without China and India b. Source: FAO, 2004a c. Source: Maddison, 2003 d. Kilocalories per capita per day e. Source: FAO, 2005 f. Source: FAO, 2003a

For the estimation of food demand for the period 2003-2030, the FAO (2003) has indicated that developing regions will show a shift towards larger food supply (kilocalories per capita per day), as well as larger consumption of specific commodities, such as cereals, sugar, oils, and animal foods, while consumption of pulses, roots, and tubers will decrease. When information from this chapter is combined with estimates of GDP growth, results can contribute to the scenario analysis of the FAO. This chapter showed that for the Asian regions, changes might occur faster than expected by the FAO projections (FAO, 2003a) and that in the coming ten years a continuation of present trends might cause a large pressure on agriculture. The European transition occurred gradually, enabling agriculture and trade to keep pace with the growth of demand. A continuation of present economic trends might cause a large pressure on the food system within ten years because changes occur much faster than present projections indicate and cause additional pressure on the global food system as well as on scarce natural resources, such as land, energy and water. These trends coincide with environmental pressure on the food system, global warming and climate change. In combination, these pressures form an enormous challenge for the system to supply sufficient food for the growing world population.

75

CChhaapptteerr 77

PPaatthhwwaayyss ttoowwaarrddss ssuussttaaiinnaabbllee ffoooodd ccoonnssuummppttiioonn ppaatttteerrnnss

Abstract Changes in food consumption patterns towards more affluent patterns is an important issue for food

security in the coming decades. This thesis focuses on the relationship between these changes and the

environmental dimension of sustainability. This chapter integrates results from preceding chapters to indicate

transition pathways towards sustainable food consumption patterns that have the most favorable characteristics

in terms of land, ‘transpirational’ water and energy use while considering nutritional and cultural constraints. It

compares differences of requirements for natural resources among food consumption patterns, in time as well as

in space. This is illustrated by the identification of long-term trends for the Netherlands and by the comparison of

resource use for the affluent pattern in the Netherlands and the average, poor food consumption pattern in

Nigeria. The time trend and comparison indicate practical information on desirable directions of change.

Foods typical for an affluent consumption pattern, beverages, fats, animal foods, fruits, and sugar have large

impacts on the use of natural resources. Especially the category of animal foods (meat, and dairy and eggs) has

a large impact. In the Netherlands in 1990, it required 66 percent of total ‘transpirational’ water, 54 percent of

agricultural land, and 45 percent of energy requirements. The requirement for natural resources results from a

combination of a specific production and consumption subsystem. When the thesis viewed long-term trends of

land use from the perspective of consumption, it used data from the Dutch production system in 1990, providing

relative results. Although actual, per capita land requirements for food in the Netherlands over the period 1950-

1990 had shown a decreasing trend, relative land requirements related to changes in per capita consumption

increased by 40 percent. For water and energy, actual ‘transpirational’ water requirements have increased by 45

percent, and actual energy requirements have almost doubled. For all resources, the effects of consumption

changes are mainly related to changes in the food categories of beverages (wine, beer, coffee) and meat (beef,

pork, poultry). For water, also larger consumption of fruits causes an increase of requirements. Affluent food

consumption patterns have much larger resource requirements than patterns from developing countries, such as

the average pattern in Nigeria. The latter shows relatively small requirements for all food categories, except for

cereals, sugar, potatoes, vegetables, and fruits that seem inversely related to requirements for the other four. Today, developing countries still have relatively small resource requirements for food. Growth potentials are

large, however. For these countries, likely pathways towards sustainable food consumption patterns may

increase resource requirements. If differences in specific requirements among food items and categories are

considered, growth can take place in a more sustainable direction. For developed countries, changes that have

occurred in the past decades provide insight into ways to promote more sustainable food consumption patterns

with smaller resource requirements. Especially the large gap between per capita supply and actual consumption

offers an opportunity to prevent resource use but it requires further research. For land, increased efficiency of

land use by increasing yield levels is an option for reduction, although this requires larger energy input. Within

food categories, substitution of foods by foods with similar functions but lower requirements is possible. This

concerns substitutions of wine by beer, coffee by tea, and soyoil by sunflower or rapeoil. In the category of meat,

resource requirements decrease in the following order: beef, pork, poultry; in the category of cereals, potatoes,

Chapter 7 __________________________________________________________________________________________

76

vegetables and fruits, requirements for staples decrease in the order of rice, pasta, potatoes. Substitution among

food categories is possible for food items of the category of meat, and dairy and eggs by protein-rich vegetal

foods, such as peas. This option shows large resource reductions.

The thesis selects important factors from the economic and social dimension of sustainability that affect the

environmental dimension through consumption. Although many relationships within the food system necessarily

had to remain unquantified, the thesis contributes in a multidisciplinary way to priority topics on prevention and

adaptation to human dimensions of environmental change. Future research should address resilience of the food

system and focus on linkages among economic, social, and environmental systems. This type of research

requires shared concepts on semantics and sustainability, collaboration efforts, commitment of scientific

disciplines and deserves a priority place on the policy agenda.

77..11 ►► IInnttrroodduuccttiioonn Sustainability or sustainable development has three dimensions, the economic, the social, and the environmental (World Commission on Environment and Development, 1987). These dimensions can also be recognized in the food system. The three dimensions include a large number of factors that all play a role in the food system. The economic dimension of the food system includes factors such as annual GDP per capita that determines the possibility to buy food, and costs and profits in agriculture; the social dimension includes factors such as health and culture that determine people‘s food choice; the environmental dimension includes factors such as the availability of natural resources, a prerequisite for agricultural production. The dimensions of the food system show a large number of relationships and interactions, among factors of the same dimension, or among factors of different dimensions. Much scientific, monodisciplinary research is available on relationships and interactions within dimensions of the food system. Results provide detailed information on parts of the food system, for example, on social factors that determine food choice (Mennell et al., 1992), or on agriculture. The quantification of all relationships, interactions, trade-offs and feed back mechanisms that exist in the food system exceeds the number of research subjects this thesis can cover by far. This means that it has to limit the number of research subjects and make strong simplifications. It focuses on one of the four global food issues that dominate food security in the coming decades, the changes in food consumption towards more affluent patterns. From this focus derive the scope of the thesis, i.e. the environmental dimension of sustainability, its emphasis, food consumption patterns, and its aim, to indicate transition pathways towards sustainable food consumption patterns. For this aim, the thesis includes elements from all three sustainability dimensions. It makes the environmental dimension operational by selecting the use of three natural resources, land, fresh water, and energy carriers. It assesses the relationship between a high productive food system and the use of land, and shows general characteristics of water requirements for food crops. It includes the economic and social dimensions by addressing the relationships between per capita income, health, emotional, and cultural factors on the one hand, and food consumption and thus the impact of a selected number of economic and social factors on natural resources on the other. The thesis shows that food consumption is an important connection between the three dimensions of sustainability of the food system. Developments towards more sustainable food consumption should consider these dimensions and the many interactions that occur among them. The aim of this chapter is twofold. First, it indicates transition pathways towards more sustainable food consumption patterns that have the most favourable characteristics in terms of land, fresh water, and energy use, while considering nutritional and cultural constraints. Second, it presents a conceptual framework of the food system, the place of this thesis in scientific knowledge, and the direction of desirable future research. The chapter integrates results from preceding chapters aimed at answering the central research question. This is illustrated by comparing resource use of an affluent pattern with a poor pattern from a developing country, and by the identification of long-term trends for the affluent food consumption pattern. Information on differences among specific food consumption patterns and their natural resource use give insight into impacts of individual foods and food categories and desirable directions of change. It is stressed that the total use of natural resources for a specific food consumption pattern depends on the combined effect of its production and consumption subsystem. This means that fruitful research on the effect of consumption subsystems on natural resources can only be carried out for a

Pathways towards sustainable food consumption patterns

__________________________________________________________________________________________

77

clearly defined production subsystem. This chapter derives information from one specific production subsystem, the Dutch system of 1990, generating relative results rather than actual resource use. In this way, the thesis contributes to a better understanding of the food consumption-environment connection, and provides practical information for more sustainable food consumption patterns. This final chapter is structured as follows. Section l compares land, ‘transpirational’ water and energy use for food consumption patterns in space and in time. Section ll applies the results obtained to provide practical information on pathways towards sustainable food consumption patterns. The general discussion presents a conceptual framework of the food system, its three sustainability dimensions, and the place of this thesis in this type of scientific research. It also provides suggestions for further research. The chapter ends with the general conclusions concerning the central research question. 77..22 ►► SSeeccttiioonn ll,, nnaattuurraall rreessoouurrccee uussee ffoorr ffoooodd ccoonnssuummppttiioonn ppaatttteerrnnss iinn ttiimmee aanndd ssppaaccee 7.2.1 Introduction Section l The three core indicators proposed in Chapter 2, land, fresh water, and energy use provide information on environmental impacts of the production of individual foods. By combining information on resource use for food items with data on per capita consumption, results show the impact of consumption patterns on natural resources. Section l compared land, ‘transpirational’ fresh water, and energy use for food consumption patterns in space and in time. For land and energy, information on requirements per unit of food is available for the Dutch production system in 1990 (Appendix A; Kramer and Moll, 1995); for water, this information is not available yet. Therefore, this section first used data from Chapter 3 and 5 to assess ‘transpirational’ water requirements for food items. Differences among resources were illustrated by comparing the use of land, ‘transpirational’ water, and energy for an affluent food consumption pattern, the Dutch pattern of 1990, and by assessing time trends for this pattern over the period 1950-1990. Differences among food consumption patterns were illustrated by comparing food supply and related resource use for the affluent consumption pattern of the Netherlands and a pattern from a poor, developing country, Nigeria. 7.2.2 Methods Section l 7.2.2.1 ‘Transpirational’ water requirements for foods This section calculated ‘transpirational’ water requirements for food items by combining the methodology for the assessment of land requirements from Chapter 3 with information on ‘transpirational’ water for crops from Chapter 5. This was done for seventeen foods from five food categories: (i) beverages (beer, wine, coffee, and tea); (ii) fats (vegetal oil); (iii) meat (beef, pork, and poultry); (iv) dairy and eggs (raw milk, butter, cheese and eggs); and (v) cereals, sugar, potatoes, vegetables and fruits. Information on food industry recipes was obtained from Kramer and Moll (1995), and information on livestock feed was adopted from Elferink (2006). The latter study assumed that for the production of one kilogram of beef, 9.6 kilogram of grass (dry matter) is needed, for one kilogram of pork, 3.43 kilogram of wheat (dry matter), and for one kilogram of poultry 2.83 kilogram of wheat (dry matter). For fruits in category five, the largest quantity of fruits consumed worldwide concerns fruits from trees, such as apples and citrus fruits (FAO, 2006). Compared to arable crops, the harvest index of fruits, i.e. the ratio of economic yield (fruits) and the total biomass production (tree biomass and fruits), is small. For the assessment of ‘transpirational’ water requirements for fruits, this chapter defined a hypothetical fruit (H-fruit) as a representative, and assumed a relatively small harvest index of 0.01. 7.2.2.2 Land, ‘transpirational’ water, and energy for the Dutch food consumption pattern One of the aims of this section was to compare the historic development of the use of land, ‘transpiratonal’ water, and energy for food consumption patterns. For the temporal analysis, it selected the development of the Dutch pattern between 1950 and 1990. This was done for two reasons. First, it concerns the development from a simple pattern in the post war period towards an affluent pattern of a western, well developed country with a relatively high, average annual GDP per capita (Maddison, 1995), and second, because of the large amount of data available for the Netherlands. The assessments were done by combining information on food supply or household consumption with data on resource requirements for individual commodities and foods. For land and water, the section combined data on supply (kg per capita per year) from the FAO (2006) for the

Chapter 7 __________________________________________________________________________________________

78

commodities and foods of the five categories mentioned above with data on land and ‘transpirational’ water requirements from Appendix A and Table 7.2. Chapter 2 showed that energy is needed in all chain links of the food production system. For the assessment of energy requirements, therefore, more detailed information on final consumption was needed. Information on household consumption (kg per household per year) for over a hundred food items was obtained from the study of Gerbens-Leenes (1999) and combined with data on energy requirements from Kramer and Moll (1995). To make results compatible, the thesis presented results in a relative way. 7.2.2.3 Long-term trends in actual land, water and energy requirements The comparison of long-term trends of land, ‘transpirational’ water, and energy requirements of the Dutch food consumption pattern based on data from the 1990 production system provided information on the impact of changes in consumption on the relative use of natural resources. Over this period, though, both the consumption and the production subsystem were changing. For example, agriculture generated larger output per unit of land. An indication of the development of decreasing land requirements per unit of food is the increasing yield of wheat from 3.3 tons per hectare in 1950 to 7.7 in 1990 (FAO, 2004a). Beside an assessment of relative resource use, this section also made an estimation of developments of actual resource requirements. For land, this section estimated actual requirements by using the increasing yield levels of wheat as an indicator. It multiplied results derived from Chapter 4 (relative land requirements) by the ratio of actual and 1990 land requirements (m2 kg-1) for wheat. For ‘transpirational’ water, changes in the biophysical production subsystem do not affect water requirements per unit of food because ‘transpirational’ water requirements are constant per unit of output. For energy, historic data on requirements for individual foods (MJ kg-1) are lacking, but information on actual, total, per capita energy requirements is available from Vringer (2005). For energy, the section first calculated time trends related to changes of annual, per capita supply using data of the production system of 1990 as input, providing relative results. Next, it evaluated the gap between relative, energy requirements for per capita food supply and actual energy requirements for total, per capita food consumption. 7.2.2.4 Resource use in a developing and in a developed country For the comparison of resource use for food in a poor, developing and in an affluent, developed country, this section selected an average pattern from Sub Saharan Africa (SSA), the Nigerian pattern, and the Dutch 1990 pattern as representatives. Nigeria has 137 million inhabitants (U.S. Census Bureau, 2004) and is the most populous country in SSA, while its average, annual GDP per capita is relatively low (Maddison, 1995). Resource requirements were calculated according to the methodology described in section 7.2.2.2. Table 7.1 shows the amounts of foods for the two patterns that formed the basis for the calculations. 7.2.3 Results and discussion Section l 7.2.3.1 ‘Transpirational’ water requirements Table 7.2 shows ‘transpirational’ water requirements for seventeen commodities and food items from five categories. Chapter 5 showed that differences among ‘transpirational’ water requirements for hypothetical crops were not substantial. When the food chain is taken into account, however, differences among individual food items arise caused by differences in the input of crops needed for the output of a unit of food. Beet sugar factories, for example, require 7.7 kg of beet to manufacture 1 kg of sugar, resulting in a relatively large ‘transpirational’ water requirement for sugar of about 300 liters per kg, the largest requirement of that nature in the food category of cereals, sugar, potatoes, vegetables, and fruits. In this category, water requirements for vegetables and fruits differed by a factor of ten. In the category of beverages, ‘transpirational’ requirements for wine and tea were twice the requirements for beer and coffee. Within the category of meat, beef had the largest ‘transpirational’ water requirement (1690 liters per kg), twice the requirement for pork (823 liters per kg), while requirements for poultry were lowest (679 liters per kg). These differences resulted from differences in conversion efficiencies among livestock (Spedding, 1988). In the category of dairy and eggs, large differences among food items occurred, ‘transpirational’ water requirements for butter, for example, were ten times larger than those for milk products. These differences were mainly due to the allocation methodology over food items produced in systems with multiple output. When requirements among categories were compared, Table 7.2 shows that the largest requirements occurred in the categories of meat, fats, and dairy and eggs; the smallest requirements in the categories of beverages, and cereals, sugar, potatoes, vegetables and fruits.


__________________________________________________________________________________________

79

Table 7.1 Food supply (kg per capita per year in Nigeria in 2000 and in the Netherlands in 1990 (Source: FAO, 2006)

Food item Supply (kg per capita per year)

Nigeria (2000)

The Netherlands (1990)

Beverages Beer 8.0 92.5 Wine 0 14.5 Coffee 0 10.4 Tea 0.2 0.8 Fats Vegetable oil and fats 14.2 17.7 Meat Beef 2.7 19.6 Pork 1.4 34.7 Poultry 1.7 18.4 Other meat 2.8 0.9 Dairy and eggs Raw milk 13.5 306.5 Butter 0.1 3.8 Eggs 3.5 11.6 Cereals, sugar, potatoes, vegetables and fruits

Cereals 150.8 70.0 Sugar 8.8 55.4 Starchy roots and tubers (e.g.potatoes) 227.0 98.7 Vegetables 52.8 78.0 Fruits 70.3 136.6

It is stressed, that the thesis took only ‘transpirational’ water use in agriculture into account, while fresh water use in the food chain fell beyond the system boundary. In this way, water requirements for foods were underestimated. A report on water use in the food industry in the United Kingdom (Whitman and Holdsworth, 1975) has made a detailed analysis of industrial water requirements. It has shown that water use in the food industry differs strongly among factories, and depends, among other factors, on factory size and efficiency. Based on that report, this thesis argued that water requirements mainly take place in the first link of the production chain, agriculture growing crops. Requirements in following links are but a minor part of total requirements. For sugar, for example, those authors have reported a requirement of 53 liters per kg of beet sugar, only 18 percent of ‘transpirational’ water requirements. For animal foods, ‘transpirational’ water requirements were relatively large because feed has to be converted into food. For the assessment, drinking water fell beyond the system boundary also causing an underestimation of water requirements. Pigs, for example, need about 0.1 liters per kg per day (Statistisches Bundesamt, 1998). Over the lifetime of a pig of 200 days (Gerbens-Leenes, 1999), pigs need 20 liters of water per kg of weight, while ‘transpirational’ water requirements were 823 liters per kg of pork. This means that neglecting drinking water caused a small underestimation of water requirements for meat.

Chapter 7 __________________________________________________________________________________________

80

Table 7.2 Overview of ‘transpirational’ water requirements for eighteen food items from five food categories, beverages, fats, meat, dairy and eggs, and cereals, potatoes, vegetables and fruits

Food item ‘Transpirational’ water requirement

(liters per kg)

Beverages

Beer 58 Wine 138 Coffee 139 Tea 56

Fats Vegetal oil 1120

Meat Beef 1690 Pork 823 Poultry 679

Dairy and eggs Raw milk 103 Butter 1116 Cheese 509 Eggs 454

Cereals, sugar, potatoes, vegetables and fruits Cereals 204 Sugar 293 Potatoes 38 Vegetables 11 Fruits 110

7.2.3.2 Resource use for the Dutch food consumption pattern in 1990 Figure 7.1a-c shows the results of the comparison of land, ‘transpirational’ water and energy requirements for the five food categories of the Dutch food supply and consumption in 1990. For the category of meat, and dairy and eggs, requirements for all three resources were relatively large. Almost 70 percent of total ‘transpirational’ water requirements, 50 percent of total land requirements, and 45 percent of total energy requirements were needed for these two food categories. For the category of cereals, sugar, potatoes, vegetables, and fruits, energy requirements were about 40 percent of the total, and ‘transpirational’ water requirements about 20 percent, while land requirements were in the same order of magnitude than for beverages, about 14 percent. The relatively large energy requirement was caused by energy use in the food chain, such as energy for transportation, production of fruits and vegetables in greenhouses, and industrial manufacture. ‘Transpirational’ water requirements for this category were also relatively large. In this category, only two food items accounted for 80 percent of all ‘transpirational’ water requirements, fruits (41 percent) and cereals (40 percent). Potatoes accounted for 10 percent, sugar for 6 percent, and vegetables for only 2 percent. The categories of beverages and fats showed the smallest requirements for ‘transpirational’ water. Land requirements for these two categories were relatively large, one third of the total land requirement was needed for beverages and fats. The fraction of energy needed for beverages was in the same order of magnitude as that for land, the fraction of energy needed for fats was much smaller than the fraction of land.


__________________________________________________________________________________________

81

7.1a

Land requirement Dutch food supply 1990

beverages13%

fats19%

meat29%

dairy and eggs25%

c,s,p,v,f14%

7.1b

'Transpirational' water requirements Dutch food supply 1990

beverages5%

fats10%

meat44%

dairy and eggs22%

c,s,p,v,f19%

Chapter 7 __________________________________________________________________________________________

82

7.1c

Energy requirement Dutch household food consumption 1990

beverages12%

fats2%

meat26%

dairy and eggs19%

c,s,p,v,f41%

Fig. 7.1a-c. Resource use for the five food categories of the Dutch 1990 food consumption pattern, beverages; fats; meat; dairy and eggs; and cereals, sugar, potatoes, vegetables, and fruits (c,s,p,v,f). Fig. 7.1a shows land requirements of supply; Fig, 7.1 b shows ‘transpirational’ water requirements of supply; Fig. 7.c shows energy requirements of household consumption. 7.2.3.3 Long-term time trends for land, water, and energy of the Dutch food consumption pattern Figure 7.2a-c shows long-term trends for land, ‘transpirational’ water, and energy requirements of the Dutch food consumption pattern over the period 1950-1990. Figure 7.2a shows the results for land requirements. The total land requirement for per capita food supply results from the combined effect of the consumption as well as the production subsystem. The consumption subsystem changed towards increased per capita supply of more and more affluent foods. The bottom line in Figure 7.2a shows the impact of these changes on land requirements. It shows that if food had been derived from the Dutch production subsystem as it existed in 1990, changes in per capita supply would have caused an increase of relative land requirements of 40 percent, an increase that mainly resulted from larger per capita supply of beverages and meat. However, over the period considered, also changes in the production subsystem occurred resulting in decreasing, actual land requirements per unit of food. This is illustrated in Figure 7.2 by the decreasing land requirement for wheat. The upper line shows that between 1950 and 1990 land requirements for wheat (m2 kg-1) more than halved. The development of actual land requirements is a combined effect of increased per capita supply of more and more affluent foods and larger output per unit of land in agriculture. The middle line in Figure 7.2a shows the actual development. Over the period 1950-1990, the combination of changes in the consumption and the production subsystem caused a decrease of actual, per capita land requirements by 40 percent. Figure 7.2b shows ‘transpirational’ water requirements for Dutch per capita food supply over the period 1950-1990. For water, changes in the consumption subsystem resulted in an increase of requirements by 45 percent. Requirements for the food category of dairy and eggs remained stable, while requirements for the other food categories increased. For the category of beverages, ‘transpirational’ water requirements increased from 1.0 to 8.5 m3 per capita per year, mainly due to larger consumption of beer and wine. For the category of fats, larger supply caused an increase of requirements from 21 to 25 m3 per capita per year. For the category of meat, ‘transpirational’ water requirements more than doubled, from 41 to 85 m3 per capita per year, mainly due to larger consumption, especially of poultry (+11 m3) and pork (+21 m3). For the category of cereals, potatoes, vegetables, and fruits, overall increase was small, but within this category, changes occurred. Requirements for cereals and potatoes decreased, requirements for sugar and potatoes remained stable, but ‘transpirational’ water requirements for fruits increased by 6 m3 per capita per year.


__________________________________________________________________________________________

83

Figure 7.2c shows trends in relative, per capita energy requirements for food supply for the five food categories, as well as actual, total, energy requirements for final, per capita consumption. Over the period 1950-1990, per capita energy requirements for supply increased by 60 percent. The main changes occurred for the categories of beverages (from 200 to 1600 MJ per capita per year), fats (from 400 to 550 MJ per capita per year), and for the category of meat (from 2000 to 4500 MJ per capita per year) For beverages, the increase of energy requirements was caused by larger consumption of coffee (+271 MJ per capita per year), wine (+400 MJ per capita per year), and beer (+650 MJ per capita per year). In the category of meat, the increase was caused by larger consumption of beef (+ 400 MJ per capita per year), poultry (+ 700 MJ per capita per year), and pork (+1300 MJ per capita per year). Figure 7.2c also shows that, during the forty-year-period, actual, per capita energy requirements for food increased even more, by 83 percent (+10 GJ per capita per year), causing a widening of the gap between energy requirements related to supply and to final consumption. The increase of this gap is understandable by assuming that more and more energy is needed in every link of the food chain, for example, in agriculture, where increasing energy inputs caused an increase of yield levels, for food conservation (e.g. cooling and freezing), and transportation. When energy for the different food categories of supply was compared with energy for total, household consumption, the impact of processes in food chains became clear. Figure 7.1c and 7.2c show energy requirements for household consumption and for per capita supply. Differences reflect additional energy requirements needed for the processing of specific commodities in food chains. Especially energy requirements for the category of cereals, potatoes, vegetables, and fruits differed, 26 percent of total supply and 41 percent of household requirements. Relative requirements for the other categories remained stable (beverages), or decreased (dairy and eggs, meat, and fats). 7.2.a

0

50

100

150

200

250

1945 1955 1965 1975 1985 1995

Inde

x (1

990

= 10

0)

....... land requirement Dutch wheat yields (1990=100)

____ land requirement per capita food supply (based on actual yields indicated by yield levels of wheat) _____ land requirement per capita food supply (based on 1990 yield levels)

Chapter 7 __________________________________________________________________________________________

84

7.2.b

0

50

100

150

200

1950 1960 1970 1980 1990

m3

per c

apita

per

yea

rc,s,p,v,fdairy and eggsmeatfatsbeverages

7.2.c

0

5000

10000

15000

20000

25000

1950 1960 1970 1980 1990

MJ

per c

apita

per

yea

r

Gap c,s,p,v,fdairy and eggsmeatfatsbeverages

Fig. 7.2a-c. Long-term trends for land, ‘transpirational’ water, and energy for the five food categories of the Dutch food consumption pattern, beverages; fats; meat; dairy and eggs; and cereals, sugar, potatoes, vegetables, and fruits (c,s,p,v,f). of the Dutch food consumption pattern over the period 1950-1990. Figure 7.2a shows land requirements for the Dutch food system. The upper line in Figure 7.2a illustrates decreasing, actual land requirements by the decreasing land requirement for wheat. The bottom line shows changes in land requirements due to changes in food consumption patterns using data from the Dutch production subsystem in 1990. The middle line in Figure 7.2a shows the combination of changes in the consumption and the production subsystem. Figure 7.2b shows ‘transpirational’ water requirements for the Dutch food consumption pattern over the period 1950-1990. Figure 7.2c shows trends in relative, per capita energy requirements for food supply for the five food categories, as well as actual, total energy requirements for per capita food consumption.


__________________________________________________________________________________________

85

When the three resources were compared, the actual trend for land offsets trends for water and energy. Where land requirements decreased by 40 percent, ‘transpirational’ water requirements increased by 45 percent, and energy requirements almost doubled. For all resources, the effect of consumption changes were mainly related to changes in the food categories of beverages (wine, beer, coffee), meat (beef, pork, poultry), and fats. For water, larger consumption of fruits caused an additional increase of requirements for the category of cereals, potatoes, vegetables and fruits. 7.2.3.4 Poor and affluent food consumption patterns Figure 7.3a, b and c show land, ‘transpirational’ water, and energy requirements for annual, per capita food supply of Nigeria and the Netherlands. Figure 7.3c also shows the additional energy required in food chains in the category ‘other’. The Nigerian pattern had relatively small requirements for all three resources for the categories of beverages, fats, meat, and dairy and eggs, while the requirement for cereals, sugar, potatoes, vegetables and fruits was relatively large. The affluent Dutch pattern had larger requirements for all food categories, except for cereals, sugar, potatoes, vegetables and fruits. Requirements for this category seem inversely related to requirements for the other four. The study of the Dutch pattern showed that foods that are typical of affluent consumption patterns, such as beverages, fats, animal foods, fruits, and sugar have substantial impacts on the use of resources. Land requirements were strongly influenced by consumption of beverages, fats, and animal foods; water requirements by the consumption of animal foods, fruits, and sugar. Energy requirements increased along with increased consumption of beverages and animal foods but also along with more affluent life styles with a demand for more ready to eat and ready to prepare foods, larger availability of foods over the year requiring more transportation, conservation and greenhouse production, and thus larger energy use in food chains. 77..33 ►► SSeeccttiioonn llll,, ppaatthhwwaayyss ttoowwaarrddss ssuussttaaiinnaabbllee ffoooodd ccoonnssuummppttiioonn ppaatttteerrnnss 7.3.1 Introduction Section ll The identification of the impact of different food items and food categories on the use of land, ‘transpirational’ water, and energy provides a tool to identify and develop pathways towards more sustainable food consumption patterns. Given a certain population size, theoretically, there are three options to make consumption patterns more sustainable, i.e. increased efficiency of production, prevention, and substitution. Options to reduce the use of energy have been elaborated in depth in Kramer (2000). This thesis, therefore, focuses on land and water use, and addresses energy to compare the natural resources. 7.3.2 Methods Section ll 7.3.2.1 Increased efficiency of production Increased efficiency of production is a strategy to improve the sustainable use of natural resources. The section addressed options to increase the efficient use of ‘transpirational’ water and land, as well as possibile trade-offs for energy use. 7.3.2.2 Prevention There is a direct relationship between consumption and the use of natural resources. Prevention of resource use of food consumption patterns can be attained in two ways, a decrease of the volume per capita of food actually consumed, or a decrease of the supply per capita of food available. The section identified options for the prevention of resource use.

Chapter 7 __________________________________________________________________________________________

86

7.3.a

Land requirement food supply

0

200

400

600

beverages fats meat dairy andeggs

c,s,p,v,f

m2 p

er c

apita

per

yea

r

NigeriaThe Netherlands

7.3.b

'Transpirational' water requirement food supply

0

20

40

60

80


c,s,p,v,f

m3 p

er c

apita

per

yea

r


7.3.c

Energy requirement food supply

0

5

10


c,s,p,v,f other

GJ

per c

apita

per

yea

r


Figure 7.3a-c. Resource requirements for the food categories beverages; fats; meat; dairy and eggs; and cereals, sugar, potatoes, vegetables, and fruits (c,s,p,v,f) of the average food consumption pattern in the Netherlands in 1990 and the average consumption pattern in Nigeria in 1998. Figure 7.3a shows land requirements of food supply, Figure 7.3.b shows ‘transpirational’ water requirements of food supply, Figure 7.3.c shows energy requirements of food supply as well as additional energy (other) required in food chains.


__________________________________________________________________________________________

87

7.3.2.3 Substitution Food items with similar functions can differ in their resource requirements. Substitution of foods with smaller resource requirements, therefore, is an option for reduction. Theoretically, substitution can take place within food categories and among categories. When nutritional constraints are taken into account, however, Table 4.1 in Chapter 4 showed that recommended daily amounts of food items are found in all food categories. From a nutritional point of view, foods from the category of meat, and dairy and eggs can be substituted by protein rich foods from the category of cereals, potatoes, vegetables and fruits (Whitney and Rolfes, 1999). The two types of substitution options are discussed in the following. Substitution within food categories Chapter 4 showed that amounts of commodities and foods that make up food consumption patterns differ strongly among affluent, western countries and result in large differences in land requirements. To define pathways towards sustainable food consumption patterns with efficient use of natural resources, this section analyzed existing food consumption patterns on the cultural scale level. In this way, it identified practical options for substitution within categories. To make these options operational, it first calculated land and ‘transpirational’ water requirements per portion size. Based on data for the Dutch production system in 1990, the section calculated smallest and largest land requirements per food category. By combining results from this analysis with information on land and water requirements per portion size, options for substitution within food categories became available. Substitution among food categories Substitution among food categories is possible for foods with similar nutritional functions. Based on the recommended amounts of foods shown in Table 4.1, only a shift from a menu that includes meat, dairy and eggs towards a vegetarian menu that contains vegetal proteins is possible when nutritional constraints are taken into account (De Wijn, 1972). This section analyzed the effect of substituting animal foods from the category of meat, and dairy and eggs by vegetal, protein rich foods, such as peas. 7.3.3 Results and discussion Section ll 7.3.3.1 Options for increased efficiency In general, water use for food is difficult to quantify because agricultural water use is site specific and determined by many factors other than crop growth. If only transpiration is considered, this flow shows a linearity with yields per unit of land. This means that water use efficiency increases with increased yield levels. For water, Chapter 5 showed that ‘transpirational’ water requirements in the first link of the food production chain, agriculture growing crops, were constant per unit of output. A more efficient use of ‘transpirational’ water, therefore, is not possible without genetic modification. For land, Figure 7.2a showed that the efficiency of the use of agricultural land in the Netherlands increased by a factor of two over the period 1950-1990. Especially in developing countries, with relatively low yield levels (FAO, 2006), a more efficient use of land, as already occurs in western countries, can decrease land requirements. In Nigeria in 1998, for example, land required for cassava was 0.9 m2 per kg, whereas IIASA projections indicate an attainable land requirement under rainfed conditions of 0.2 m2 per kg (Gerbens-Leenes, 2006). Increased efficiency of land use, however, goes along with larger inputs of, for example, fertilizer, and thus with larger energy use. Although energy is used in all links of a production chain, trade-offs between land and energy have to be taken into account when strategies are developed to improve the efficient use of land. Table 2.1 in Chapter 2 and Appendix A show that the production of one kg of French beans in the open air requires 1 MJ of energy and 0.9 m2 of land, whereas the production in a greenhouse requires 26 MJ and 0.3 m2. This trade-off shows that the different aspects of sustainability make it difficult to formulate general rules for more sustainable food systems and that always specific constraints have to be taken into account. Section 7.2.3 showed that energy use in food chains mainly concerns the food category of cereals, potatoes, vegetables, and fruits. A more efficient use of energy should therefore focus on this food category. This points to the need for future research in order to analyze trade-offs between land and energy, as well as to indicate pathways towards more efficient energy use in production systems.

Chapter 7 __________________________________________________________________________________________

88

7.3.3.2 Options for prevention Chapter 6 showed that, in general, per capita food actually consumed (intake of kilocalories per capita per day) is rather constant and in line with physiological requirements. Eating less is therefore no sensible option to decrease resource use. Table 4.2 in Chapter 4 showed the development of per capita supply in the Netherlands between 1950 and 1990. Especially the increase in the food category of beverages is large. Drinking less, especially beverages with large consumption or relatively large resource requirements, such as beer, wine and coffee, therefore, could be an option to prevent resource use. Chapter 6 also showed that the gap between per capita supply and actual consumption increases along with increasing GDP per capita, especially for low income categories. For developed countries, the thesis found a ratio of supply over final consumption of 1.6, indicating that a large part of available food is wasted in the food chain and not eaten at all. Prevention of food waste could reduce resource use. An analysis of these wastes fell outside the scope of this thesis, but, considering the magnitude of the gap between average per capita supply and final consumption, this requires further research. 7.3.3.3 Options for substitution within food categories Table 7.2 shows serving size, and land and ‘transpirational’ water requirements per serving for food items from five food categories. For food items from the same category, large apparent differences in resource requirements are providing a tool for reduction. Table 7.3 shows smallest and largest land requirements for the five food categories found for existing food consumption patterns, as well as the countries showing these requirements. When results per category were summed, a difference of a factor of two occurred. By combining information from Table 4.3, 7.2, and 7.3, the section identified options for substitution within food categories that are discussed in the following. Table 7.2 Land and water requirements per serving size for food items from five food categories Food item type Serving sizea. Land requirement

(m2 per serving) ‘Transpirational’ water requirement (liters per serving)

Beverages Beer Glass 200 ml 0.10 11.6 Wine Glass 100 ml 0.15 13.8 Coffee Cup 125 ml 0.10 17.4 Tea Cup 125 ml 0.04 7.0 Fats Soyoil 1 kg 26.00 1120.0 Sunfloweroil 1 kg 13.00 - Rapeoil 1 kg 7.00 - Meat Beef Serving 100 g 1.94 169.0 Pork Serving 100 g 1.20 82.3 Poultry Serving 100 g 1.02 67.9 Dairy and eggs Whole milk Glass 200 ml 0.24 20.6 Eggs Serving 50 g 0.18 22.7 Cereals, sugar, potatoes, vegetables and fruits

Rice Serving 80 g 0.27 16.3 Pasta Serving 80 g 0.15 - Potatoes Serving 200 g 0.04 7.6 Peas Serving 100 g 0.20 12.1 French beans (open air) Serving 200 g 0.18 2.2 French beans (greenhouse)

Serving 200 g 0.06 2.2

Carrots (open air) Serving 200 g 0.04 2.2 Apples Serving 200 g 0.08 22.0 Strawberries Serving 200 g 0.12 2.2 Cherries Serving 200 g 0.84 22.0

a. Source: Voedingscentrum, 1998a


__________________________________________________________________________________________

89

Table 7.3 Smallest and largest land requirement per food category for affluent food consumption patterns of the European Union and of the United States Land requirements food categories (m2 per capita per year) Beverages Fats Meat Dairy and

eggs C,s,p,v,fa. Total

Greece 74 Germany 438 Sweden 446 United States

139 Finland 193 1290

The Netherlands

237 Denmark 901 United States

924 France 512 Greece 391 2965

a. Cereals, sugar, potatoes, vegetables, and fruits ► Beverages In the category of beverages, the largest difference in land requirements among the countries investigated occurred between the consumption patterns in Greece and in the Netherlands. In Greece, consumption of wine is larger, in the Netherlands, especially the large consumption of coffee and beer resulted in large per capita land requirements. Table 7.2 and Appendix A show that within this category, substitution to reduce natural resource use is possible. For example, substitution of coffe by tea, or wine by beer. ► Fats In the category of fats, the largest difference in land requirements occurred etween the consumption pattern in Germany and in Denmark. This differnce was caused by different amounts of consumption. Table 7.2 shows that options to reduce land requirements are the substitution of soyoil, a basic ingredient of vegetal oils and margarines, by rape or sunflower oil. For water, sufficient data were lacking. ► Meat In the category of meat, the largest difference in land requirements occurred between the average consumption pattern in Sweden and in the United States, a difference mainly caused by large differences in consumption. In this category, differences among specific land and water requirements provide options for reduction. Requirements of land and ‘transpirational’ water for meat decrease in the following order: beef, pork, poultry. It is stressed, however, that agricultural land can have different qualities. This is a relevant factor when substitution options are considered. Cattle producing beef, for example, are often fed with grass derived from pastures that, in general, have a lower quality than cropland, producing feed for pigs and poultry (Gerbens-Leenes et al., 2002). ► Dairy and eggs In the category of dairy and eggs, the largest difference in land requirements occurred between the consumption pattern in France and in the United States, a difference caused by the amounts of consumption per capita. This category offers no substitution options for reduction of resource use because requirements for milk and eggs are in the same order of magnitude. ► Cereals, potatoes, vegetables, and fruits In the category of cereals, potatoes, vegetables, and fruits, the largest difference in land requirements occurred between the consumption pattern in Finland and in Greece. The calculations were based on average numbers, however, because detailed information on a household level was not available. The data in Table 7.2 and Appendix A show that within this category, substitution is possible for foods with the same function. For example, for vegetables and fruits, Appendix A shows that differences in land requirements occur. French beans, for example, have 50 percent larger land requirements than carrots. Substitution of vegetables or fruits with relatively large land requirements by foods showing lower requirements, however, is not in line with nutritional constraints promoting a high diversity of consumption (Whitney and Rolfes, 1999). Staples have decreasing requirements in the following order: rice, pasta and potatoes providing options for reduction. When cultural constraints are also taken into account, for example, Table 4.2 in Chapter 4 shows that the Irish consumption pattern includes more potatoes than the Italian one, where people prefer cereals, the options for substitution are likely to be few. 7.3.3.4 Options for substitution among food categories Table 7.2 shows that substitution of food items from the category of meat, and dairy and eggs by high protein foods from the category of cereals, sugar, vegetables, potatoes and fruits, such as peas, cause a substantial reduction of the use of land as well as ‘transpirational’ water. The substitution of 100 grams of beef by 100 grams of peas, for example, causes a large reduction of the use of land and ‘transpirational’ water, in which a serving of peas needs only 10 percent of the land and 7 percent of the ‘transpirational’ water requirements of a serving of beef. It is stressed, however, that the comparison did not take different land qualities into account in which peas are grown using high quality cropland and beef can be produced using lower quality pasture land.

Chapter 7 __________________________________________________________________________________________

90

77..44 ►► GGeenneerraall ddiissccuussssiioonn Chapter 1, the introduction of this thesis, presented the two main components of the food system, the production and consumption subsystem that, together, determine the requirement for natural resources. The conceptual framework, as presented in Figure 1.1, formed the starting point for the research presented in this thesis. Retrospectively, the study of the food system has shown that the initial conceptual framework is a strong simplification. The understanding of the system has gradually developed towards the conceptual framework presented in Figure 7.4. It shows the three sustainability dimensions of the food system, the economic, the social, and the environmental dimension, as well as the two main components, agricultural production and the food industry, and human consumption. The three dimensions include many factors that influence the system. Apart from the food system, these factors can also be part of other systems. Many factors that play a role in the food system are interrelated, within as well as among the three dimensions. The food system is an integrated system, highly complex, showing besides linear relationships also non linear relationships, multiple equilibria, as well as thresholds. Figure 7.4 shows the principle factors this thesis has taken into account. It is stressed that many factors, represented in Figure 7.4 by open spaces, were not addressed and require further identification and research. To answer the central research question, the design and development of transition pathways towards sustainable food consumption patterns, the thesis mainly focused on the central role of consumption in the food system, as well as on the different requirements of food consumption patterns for natural resources, while the impact of the production subsystem received little attention. The thesis has opted for a multidisciplinary approach to study the consumption-environment relationship. It has selected important factors from the economic and social dimension of sustainability that affect the environmental dimension through consumption. Figure 7.4 shows that, basically, there are four types of relationships in the food system: relationships within the same sustainability dimension, relationships between factors of a specific dimension and the production and consumption subsystem, a relationship between the production and consumption subsystem, and relationships across dimensions. This thesis has addressed some relationships, but many remained unquantified. The environmental dimension, for example, includes many more factors than the three natural resources addressed in here. Factors that, either on the same scale level, or across levels, play a role in the environmental sustainability of the food system. Examples are biodiversity, land quality or nitrogen pollution from agriculture. The thesis has quantified the relationship between annual GDP per capita, and health, emotional, and cultural food requirements on the one hand, and food consumption patterns on the other. This has given insight into the effect of consumption on natural resource use and has shown reduction strategies. However, the conceptual framework presented in Figure 7.4 makes clear that much still remains unknown. The identification of indicators to assess the environmental dimension in a uniform way, for example, is still an important challenge. Another issue concerns the trade offs and interplays between and among factors, for example, the trade off that occurs between land and energy use in different agricultural production systems. An important issue is also existence of constraints from the three dimensions on the production and consumption subsystem. Agricultural production, for example, is limited within environmental constraints, such as the availability of arable land of sufficient quality, and the daily availability of fresh water. Although the output of the production subsystem has consistently shown an increase over time, physical limits to growth do exist, so that eventually an equilibrium between production and consumption will be reached. Production is also bound to economic constraints, such as the availability of capital and the need to make sufficient profit. Relationships across sustainability dimensions are, for example, the relationship between annual GDP per capita and population size, or between the need for energy crops in the social system and the availability of land in the environmental system. Here, the need to produce energy through arable crops is conflicting with the food producing function of arable land areas from the environmental dimension. Two additional factors of high importance are knowledge in the social system and technology in the economic system. Knowledge plays an important role, for individual human beings as well as for society as a whole, and stimulates the development of technology necessary for sustainable development. By selecting some priority topics, this thesis has studied several relationships of the food system in a multidisciplinary way, and has contributed to the challenges for research on prevention and adaptation to human dimensions of environmental change. Other priority topics for the near future are, for example, robustness of the model, availability of data, uncertainty, spatial specificity, problem interdependencies and stakeholder involvement. Resilience of the food system, i.e. options to adapt to global change, is an important topic for research in the near future and should focus on linkages


__________________________________________________________________________________________

91

among economic, social and environmental systems. This type of research not only requires shared concepts on semantics and sustainability, but also collaboration efforts and commitment of various scientific disciplines, as well as a priority place on the policy agenda. Fig. 7.4. Conceptual framework of the food system illustrating the three sustainability dimensions of the food system, the economic, the social, and the environmental dimension, as well as the two main components, agricultural production and the food industry, and human consumption. The three dimensions include many factors that influence the system. Apart from the food system, these factors can also be part of other systems. Many factors that play a role in the food system are interrelated, within as well as among the three dimensions. 77..55 ►► GGeenneerraall ccoonncclluussiioonnss In order to contribute to a better understanding of the consumption-environment relationship, this thesis applied a systems approach and expressed environmental sustainability in food systems on a global scale level in terms of land, ‘transpirational’ water and energy requirements. The design and development of measuring methods for environmental sustainability in food production systems focused on chains and webs. Insight into sustainability aspects of food production can lead towards more sustainable consumption patterns. The adoption of only three sustainability indicators, land, ‘transpirational’ water, and energy, in combination with a systems approach contributed to an increased transparency of the food system. In this way, the thesis addressed the main functions of the environment: the source, sink, life support, and human health and welfare function, and addressed some important global sustainability issues, such as climate change, food security, and fresh water

Social system

- Health - Knowledge - Governance - Culture - Emotion - Population size - Consumption pattern -

Environmental system

- Energy - Land - Water - Biodiversity - Nitrogen pollution - -

Agricultural production and food

industry

Human consumption

TOTAL

RESOURCE REQUIREMENT FOR

FOOD

Economic system

- Costs and profits - Technology - Annual GDP per capita - -

Chapter 7 __________________________________________________________________________________________

92

availability. Results showed that food production needs land and ‘transpirational’ water mainly in the first chain link, agriculture, energy in all other links. However, the thesis observed trade offs among the three indicators implying that a reduction of the use of one resource might require larger input of another. This means that resources cannot be minimized but should rather be optimized within existing environmental constraints. The need for temporal and regional adaptations needs further research. The effect of food consumption patterns on natural resources results from the combined effect of specific requirements of individual foods for natural resources in combination with amounts of foods consumed. Specific requirements for foods produced in the same system vary, leading to variation of resource requirements among food consumption patterns, in time as well as among countries. Especially affluent food consumption patterns from developed countries have much larger requirements for agricultural land, ‘transpirational’ water and energy than poor patterns from developing countries. Moreover, there are also substantial differences in requirements among patterns from western countries, because some food categories have larger impact on resources than others. For the affluent Dutch pattern of 1990, the category of animal foods (meat, and dairy and eggs) required 66 percent of total ‘transpirational’ water, 54 percent of agricultural land, and 45 percent of energy requirements. Beverages and fats required relatively large amounts of land, but less water and even less energy. The food category of cereals, potatoes vegetables and fruits required relatively large amounts of energy, about 40 percent, which can be attributed to energy needed in chains, such as for the manufacture, transportation and conservation of foods. Long term trends for the Netherlands showed that the pressure of consumption on land, ‘transpirational’ water and energy has increased substantially over a forty year period. Especially increased consumption in the food category of meat caused large changes in land, ‘transpirational’ water, and energy requirements, additional effects were caused by increased consumption of beverages. For energy, the increased energy use in chains was apparent. The answer on the central research question, what are desirable transition pathways towards sustainable food consumption patterns that have the most favorable characteristics in terms of land, fresh water, and energy use while considering nutritional and cultural constraints, derives from changes that have occurred in the past decades in developed countries, as well as differences in consumption among countries. These changes and differences provide insight into ways to promote sustainable food consumption patterns. Tools are increased efficiency, prevention and substitution. The thesis shows that developing countries still have relatively small natural resource requirements for food, mainly needed for the category of cereals, potatoes, vegetables and fruits. Growth potentials are large, though. In these countries, however, pathways towards sustainable food consumption patterns can not consider a reduction of resource requirements, but should first address the economic and social aspects of sustainability, including nutritional and cultural constraints. In developed countries, pathways towards sustainable food consumption patterns can reduce resource requirements. Especially the large gap occurring between average food supply per capita and actual consumption is an option to prevent resource use but requires further research. For land, increased efficiency of land use by increasing yield levels is a reduction option, although this would require larger energy input. Within food categories, substitution of foods by foods with similar functions but smaller resource requirements is possible. These substitutions are, for example, wine by beer; coffee by tea; and soyoil by sunflower or rapeoil. In the category of meat, requirements decrease in the following order: beef, pork, poultry; in the category of cereals, potatoes, vegetables and fruits, requirements for staples decrease in the order of rice, pasta, potatoes. Substition among food categories is possible for food items of the category of meat, and dairy and eggs by protein rich vegetal foods, such as peas. This option shows large resource reductions. Resilience of the food system, i.e. options to adapt to global change, is an important topic for research in the near future. It should focus on linkages among economic, social and environmental systems and address robustness of models, availability of data, uncertainty, spatial specificity, problem interdependencies and stakeholder involvement. This type of multidisciplinary research requires shared concepts on semantics and sustainability, collaboration efforts, commitment of various scientific disciplines, as well as a high priority on the policy agenda. In this way, research can contribute to an adequately nourished global population, which is an important constraint for sustainable development.

93

RReeffeerreenncceess

Achtienribbe, G.E., 1993. Household water consumption in the Netherlands. J Water SRT-Aqua 42, 347-350.

Agricultural Research Service, 1999. U.D.o.A.U. Continuing Survey of Food Intakes by Individuals. http://www.Barc.usda.gov/bhrnc/foodsurvey/pdf/Region.pdf/.

Amat, C., Curonisy, P., 1981. La alimentación en el Perú. Centro de Investigación Universidad del Pacifico, Lima, Perú.

Amthor, 1984. The role of maintenance respiration in plant growth. Plant Cell Environ., 7: 561-569.

Andersson, K., 2000. LCA of food products and prodcution systems. International Journal of LCA, 5(4): 239-248.

Arkebauer, T.J., Weiss, A., Sinclair, T.R., and Blum, A., 1994. In defense of radiation use efficiency: a response to Demetriades-Shah et al. (1992). Agricultural and forest meteorology, 68: 221-227.

Asociación Demográfica Salvadorĕna (ADS), Ministerio de Salud Pública y Asistencia Social (MSPAS), Instituto de Nutrición de Centro América y Panamá (INCAP), 1990. Evaluación de la Situación Alimentaria Nutricional en El Salvador, ESA NES-88, El Salvador.

Avila, A., Shamah, T., Chavez, A., 1997. Enquestas de Alimentación y Nutrición en el Medio Rural, 1996. Resultados por entidad. INNSZ, DEDESOL, DIF, SSA, Golernos de los Estados, IMSS, INI, Unicef, Mexico.

Azoulay, G., 1998. Enjeux de la sécurité alimentaire mondiale. Cahiers Agriculture, 7: 433-439.

Bakker, Th., 1984. Horizonten van zelfvoorziening. Mogelijkheden voor de voedselvoorziening van Nederland in autarkische omstandigheden. Landbouw-Economisch Instituut, Den Haag.

Bansal, P., Kilbourne, W.E., 2001. The ecologically sustainable retailer. Retailing and consumer services, 8: 139-146.

Bockstaller, C., Girardin, P., Werf van der, H.M.G., 1997. Use of agro-ecological indicators for the evaluation of farming systems. European Journal of Agronomy, 7: 261-270.

Boer, M. de., Wijers, G.J., Aartsen, J.J.van, Jorritsma-Lebbink, A., Vermeend, W.A.F.G., Pronk, J.P., 1998. Nationaal Milieubeleidsplan 3. ‘s-Gravenhage, The Netherlands.

Bolhuis, J., Klein de, A.J., Pronk, A., 1995. Jaarstatistiek van de veevoeders 1990/’91 en 1991/’92. Landbouw-Economisch Instituut (LEI-DLO), ’s-Gravenhage.

Boom-Binkhorst van der, F.H., Winkelman, M.L.J., Lith van, A., von Lossonczy von Losoncz, T.O., Amesz, M.F., Schure-Remijn, P.J.M., 1997. Voedingsmiddelen en dranken in het Nederlandse voedingspatroon. In: Stafleu, A., Veen, J.M., Vredebregt-Lagas, W.H. (Eds.), Mens en Voeding. Nijgh en van Ditmar, Baarn, the Netherlands, pp. 233-234.

References ___________________________________________________________________________________

94

Boons-Prins, E.R., Koning de, G.H.J., Diepen van, C.A., Penning de Vries, F.W.T., 1993. Crop specific simulation parameters for yield forecasting across the European Community. CABO-DLO/SC-DLO. Simulation Reports CABO-TT 32. Wageningen, the Netherlands.

Boudouropoulos, I.D., Arvanitoyannis, I.S., 1999. Current state and advances in the implementation of ISO 14000 by the food industry. Comparison of ISO 14000 to ISO 9000 to other environmental programs. Trends in Food Science & Technology, 9: 395-408.

Bouma, J., Batjes, N.H., Groot, J.J.R., 1998a. Exploring land quality effects on world food supply. Geoderma, 86: 43-59.

Bouma, J., Varallyay, G., Batjes, N.H., 1998b. Principal land use changes anticipated in Europe. Agr. Ecosyst. Environ. 67, 103-119.

Bouwman, M., 2000. Tracking Transport Systems, an environmental perspective on passenger transport modes. Center for Energy and Environmental Studies (IVEM), Groningen, the Netherlands.

Braun von, J., 1988. Effects of technological change in agriculture on food consumption and nutrition: rice in a West African Setting. World Development 9, 1083-1098.

Braun von, J., Paulino, L., 1990. Food in sub-Saharan Africa, Trends and policy challenges for the 1990s. Food Policy, 505-517.

Britos, S., Scacchia, S., 1998. Disponibilidad y consumo de alimentos en Argentina. Escuela de Nutrición. Universidad Nacional de Buenos Aires, Argentina.

Brown, L.R., Renner, M., Flavin, C., Starke, L. (Eds.), 1997. Vital Signs 1997 1998, Worldwatch Institute, London, pp. 26-27.

Buishand, T.A., Velds, C.A.1980. Neerslag en verdamping. Klimaat van Nederland. Koninklijk Nederlands Meteorologisch Instituut, De Bildt, the Netherlands.

Bursztijn, P.G., 1985. A diet survey in Zimbabwe. Human Nutr. Appl. Nutr., 39 (5), 376-88.

Catsberg,C.M.E. and Kempen-van Dommelen,G.J.M., 1997. Levensmiddelenleer. Uitgeverij Intro, Baarn, The Netherlands.

Carlsson-Kanyama, A., 1997. Weighted average source points and distances for consumption origin-tools for environmental impact analysis? Ecological Economics, 23: 15-23.

Carlsson-Kanyama, A., 1998. Climate change and dietary choices - how can emissions of greenhouse gases from food consumption be reduced? Food Policy, 23(3/4): 277-293.

Carlsson-Kanyama, A., Faist, M., 2000. Energy use in the food sector: a date survey. 291. AFR-report 291. Forskningsgruppen för Miljöstrategiska Studier. Stockholm, Sweden.

Carson, R., 1962. Silent spring.

Centraal bureau voor de statistiek (CBS), 1980. Maandstatistieken van de prijzen. CBS, Voorburg, the Netherlands.


References

___________________________________________________________________________________

95

Centraal bureau voor de statistiek (CBS), 1990. Statistisch jaarboek 1990. CBS, Voorburg, the Netherlands.


Centraal bureau voor de statistiek (CBS), 1993. Statistisch jaarboek 1993. CBS, Voorburg, the Netherlands.

Centraal bureau voor de statistiek (CBS), hoofdafdeling statistieken van inkomen en consumptie, 1994. Gebruikershandboek budgetonderzoek 1992. CBS, Voorburg, the Netherlands.

Centraal Bureau voor de Statistiek (CBS), 2002. Statistisch Jaarboek 2002. Voorburg, the Netherlands.

Centraal Veevoederbureau, 1997. Voedernormen landbouwhuisdieren en voederwaarde veevoeders 22. Cvb-reeks.

Conforti, P., Giampietro, M., 1997. Fossil energy use in agriculture, an international comparison. Agriculture, Ecosystems & Environment 65, 231-243.

Costanza, R., Daly, H.E., 1992. Natural capital and sustainable development. Conserv. Biol., 6: 37-46.

Daly, H.E., 1990. Toward some operational principles of sustainable development. Ecological Economics, 2: 1-6.

Department of Census and Statistics, 1993. Household Income and Expenditure Survey 1990/91, Final Report. Department of Census and Statistics, Ministry of Policy Planning and Implementation, Sri Lanka.

Department of Statistics (DOS), 1997. Household Income and Expenditure Survey. Amman, Jordan.

Dieren van, W., 1995. Taking nature into account: a report to the Club of Rome. Springer-Verlag, New York.

Djazayery, A., Samimi, B., 1996. Food consumption and energy intake patterns in the rural and urban areas of Iran, 1983-1992. Agricultural Economics and Development, 4th Year: 218-248.

Drumwright, M., 1994. Socially responsible organizational buying: environmental concerns as a non-economic buying criteria. Journal of Marketing, 58: 1-19.

Dutilh, C.E., Kramer, K.J., 2000. Energy consumption in the food chain. Comparing alternative options in food production and consumption. Ambio, 29(2): 98-101.

Elferink, E., 2006. Center for Energy and Environmental Studies, University of Groningen, the Netherlands. Personal communication.

Ekins, P., Folke, C., Groot de, R., 2003. Identifying critical natural capital. Ecological Economics, 00: 1-5.

Ekins, P., Simon, S., 2003. An illustrative application of the CRITINC framework to the UK. Ecological Economics, 00: 1-21.

References ___________________________________________________________________________________

96

Ekkes, J.J., Bakker, J.P., Schoonhoven van, A., 1994. Kwantitatieve informatie voor de glastuinbouw 1994-1995. Informatie en Kennis Centrum Akker- en Tuinbouw, Afdeling Glasgroente en Bloemisterij, Aalsmeer/Naaldwijk, the Netherlands.

Elkington, J., 1994. Towards the sustainable corporation: win-win-win business strategies for sustainable development. California Management Review, 36(2): 90-100.

Engelenburg van, B.C.W., Rossum van, T.F.M., Blok, K., Vringer, K., 1994. Calculating the energy requirements of household purchases, a practical step by step method. Energy Policy 22, 648-656.

European Community, 1993. EUR-Lex: Community legislation in force. Official journal NO.C 138 493Y0517(01).

Eurostat, 1993. Basisstatistieken van de gemeenschap. Luxemburg, Bureau voor officiële publicaties der Europese Gemeenschappen.

Falkenmark, M., 1989a. Comparative hydrology - a new concept. In: M. Falkenmark and T. Chapman (Editors), Comparative hydrology. An ecological approach to land and water resources. Unesco, Paris, France, pp. 10-42.

Falkenmark, M., 1989b. Water scarcity and food production. In: D.Pimentel and C.W.Hall (Eds.), Food and natural resources. Academic Press, San Diego (CA), pp. 164-191.

Falkenmark, M., 1997. Meeting water requirements of an expanding world population. In: D.J. Greenland, P.J. Gregory, and P.H. Nye (Editors), Land Resources: On the edge of the Mathusian precipice? pp. 929-936.

Fernandes, E.C.M., Nair, P.K.R., 1986. An evaluation of the structure and function of tropical homegardens. Agricultural systems 21, 279-310.

Fogel, R. W., L. A. Helmchen, 2002. Economic and Technological Development and Their Relationships to Body Size and Productivity. The nutrition transition, diet and disease in the developing world. Ed. B. Caballero and B. M. Popkin. Food Science and Technology International Series. Ed. S. L. Taylor. London, San Diego: Academic Press, 9-24.

Food and Agricultural Organization of the United Nations (FAO), 1999. FAO Yearbook Production. FAO Statistics Series, vol.156, Rome.

Food and Agricultural Organisation of the United Nations (FAO), 1999. http://www.fao.org/



Food and Agricultural Organisation of the United Nations (FAO), 2003a. World agriculture : towards 2015/2030. An FAO perspective. Bruinsma, J. Earthscan Publications Ltd, London.

Food and Agricultural Organisation of the United Nations (FAO), 2003b. http://www.fao.org/

Food and Agricultural Organisation of the United Nations (FAO), 2004a. Food Balance Sheet 2004. http://www.fao.org Oktober 2004.

Food and Agricultural Organisation of the United Nations (FAO), 2004b. L’état de l’insécurité alimentaire 2001 dans le monde. Organisation des Nations Unies pour l’alimentation et l’agriculture, Rome, Italy. http://www.fao.org January 17th 2006.

References

___________________________________________________________________________________

97

Food and Agricultural Organisation of the United Nations (FAO), 2005. Departement Economique et Social, Alimentation et nutrition. Profiles nutritionnels par pays. Mali. http://www.fao.org/es/nutrition/mli-f.stm. January 17th 2006.


Food and Agricultural Organisation of the United Nations (FAO), Smith, M., 1992. Cropwat. A computerprogram for irrigation planning and management. FAO irrigation and drainage paper 46, Rome.

Food and Nutrition Research Institute of the Department of Science and Technology (FNRI- DOST) of the Philippines, 2000. National Survey of 1993: Final Results.

Freire, W., 1988. Diagnóstico de la situación alimentaria y nutricional y de salud de la población ecuatoriana menor de cinco años – DANS -1986. CONADE, MSP, Quito, Equador.

Galeazzi, M.A.M., Falchoni, P. Jnr., 1998. Inquérito de Consumo Alimentar da Area Metropolitana de Brasíla-Relatório. Técnico-Secretaria de Saúde de Brasília.

Ge, K., Zhai, F., Yan, H., 1996. The dietary and nutritional status of Chinese population. 3td

National Nutrition Survey, 1992. People’s Med. Pub. House, Beijing, China.

Gerbens-Leenes, P.W., 1999. Indirect ruimte- en energiebeslag van de Nederlandse voedselconsumptie. Center for Energy and Environmental Studies (IVEM). IVEM-onderzoeksrapport 102. Groningen, the Netherlands.

Gerbens-Leenes, P.W., 2000. Groen Kookboek. Milieubewust koken met een laag energie- en landgebruik. 103. Center for Energy and Environmental Studies (IVEM). IVEM-onderzoeksrapport 103. Groningen, the Netherlands.

Gerbens-Leenes, P.W., Nonhebel, S., 2002. Consumption patterns and their effects on land required for food. Ecological Economics, 42: 185-199.

Gerbens-Leenes, P.W., Nonhebel, S., Ivens, W.P.M.F., 2002. A method to determine land requirements relating to food consumption patterns. Agriculture, Ecosystems & Environment, 90: 47-58.

Gerbens-Leenes, P.W., Moll, H.C., Schoot UIterkamp, A.J.M., 2003. Design and development of a measuring method for environmental sustainability in food production systems. Ecological Economics 46, 231-48.

Gerbens-Leenes, P.W., Nonhebel, S., 2004. Critical water requirements for food, methodology and policy consequences for food security. Food Policy, 29, 547-64.

Gerbens-Leenes, P.W., Nonhebel, S., 2005. Food and land use. The influence of consumption patterns on the use of natural resources. Appetite, 45: 24-31.

Giupponi, C., 1998. Environmental evaluation of alternative cropping systems with impact indices of pollution. European Journal of Agronomy, 8: 71-82.

Global Reporting Initiative, 2000. Sustainability Reporting Guidelines on Economic, Environmental and Social Performance. Interim Secretariat Global Reporting Initiative, Boston, USA.

References ___________________________________________________________________________________

98

Goudriaan, J., 1982. Potential production processes. In: F.W.T. Penning de Vries and H.H. Laar van (Eds.), Simulation of plant growth and crop production. Centre for Agricultural Publishing and Documentation, Wageningen, the Netherlands, pp. 98-113.

Goudriaan, J., Groot, J.J.R., Uithol, P.W.J., 2001. Productivity of Agro-ecosystems. In: Terrestrial Global Productivity. Academic Press, pp. 301-313.

Grigg, D., 1994. Income, industrialization and food consumption. Tijdschrift voor economische en sociale geografie, 85: 3-14.

Grigg, D., 1995. The nutrition transition in Western Europe. Journal of Historical Geography 22 (1), 247-61.

Groot, J.J.R., Penning de Vries, F.W.T., Uithol, P.W.J., 1998. Food supply capacity study at global scale. Nutrient Cycling in Agroecosystems, 50: 181-189.

Groenendaal, B., 1996. Autarkisch Nederland: een onderzoek naar verschillende voedselpakketten en het bijbehorend energieverbruik. Center for Energy and Environmental Studies (IVEM), IVEM-doctoraalverslag 47, Groningen, the Netherlands.

Habekotté, B., 1996. Winter oilseed rape. Analysis of yield formation and crop type design for higher yield potential. Landbouwuniversiteit Wageningen, the Netherlands.

Habekotté, B., 1997. Identification of strong and weak yield determining components of winter oilseed rape compared with winter wheat. European Journal of Agronomy, 7: 315-321.

Hassanyn, A.S., 2000. Food Consumption Pattern and Nutrients Intake Among Different Population Groups in Egypt. Final Report (Part 1), Nutrition Institute, Egypt, WHO/EMRO.

Halberg, N., 1999. Indicators of resource use and environmental impact for use in a decision aid for Danish livestock farmers. Agriculture, Ecosystems & Environment, 76: 17-30.

Hall, J., 2000. Environmental supply chain dynamics. Journal of Cleaner Production, 8(6): 455-471.

Hansen, J.W., 1996. Is Agricultural Sustainability a Useful Concept? Agricultural Systems, 50: 117-143.

Harris, J.M., 1996. World agricultural futures: regional sustainability and ecological limits. Ecological Economics, 17: 95-115.

Heijungs, R., Guinée, J.B., Huppes, G., Lankreijer, R.M., Udo de Haes, H.A., Wegener, A., Ansems, A.M.M., Eggels, P.G., Van Duin, R., De Goede, H.P., 1992. Milieugerichte levenscyclusanalyses van produkten - Handleiding. The Netherlands, Centrum voor Milieukunde Leiden (CML), Nederlandse Organisatie voor Toegepast Natuurwetenschappelijk Onderzoek (TNO), Apeldoorn en Bureau Brand- en Grondstoffen (B&G), the Netherlands.

Hundd, Moh, 1997. Food consumption survey in 7 provinces, Project Report. Hacettepe University, Department of Nutrition and Dietetics, Ministery of Health. Ankara, Turkey.

Informatie- en Kenniscentrum voor de Akkerbouw en de Groenteteelt in de Vollegrond (IKC-AG), 1993. Kwantitatieve Informatie voor de Akkerbouw en de Groenteteelt in de Vollegrond, Bedrijfssynthese 1993-1994. Belser, Lelystad, the Netherlands.

INNSZ, 1990. Encuesta Nacional de Alimentación en el Medio Rural ENAL 1989. INCMNSZ, México.

References

___________________________________________________________________________________

99

Institute of Nutrition and Food Hygiene (INFH), 1985. Summary Report of the 2nd National Nutrition Survey in 1982. Institute of Nutrition and Food Hygiene, Beying, China.

Instituto de Nutrición de Centro América y Panamá (INCAP), Oficina de Investigaciones Internationales de Salud, Ministerio de Salud Pública y Asistencia Social (MSPAS), 2000. Evaluación nutricional de El Salvador 1969.

IPCC (Intergovernmental Panel on Climate Change), 2001. Climate Change 2001: Impacts, Adaptation, and Vulnerability- A Report of Working Group II of the IPCC.

IPCC (Intergovernmental Panel on Climate Change), 1996. Climate change 1995. The science of climate change. Contribution of working group I to the second assessment. Report of the Intergovernmental Panel on Climate Change. Australia, Cambridge University Press.

ISO World, 2000. The number of ISO 14001/EMAS certification registration of the world. http://www.ecology.or.jp/

Ivens, W.P.M.F., Dankert, G., Eng van der, P.J., Faber, D.C., van Keulen, H., Klaver, W., Lövenstein, H.M., Makken, F., Rabbinge, R., Schoffelen, E.P.L.M., Drijver-de Haas, J.S., 1992. World Food Production. Open universiteit, Heerlen, the Netherlands.

Jahan, Hossein, 1998. Malnutrition in Bangladesh: Bangladesh National Nutrition Survey, 1995-96. Institute of Nutrition and Food Science, Dhaka University, Bangladesh.

Jobse-van Putten, J., 1995. Eenvoudig maar voedzaam: cultuurgeschiedenis van de dagelijkse maaltijd in Nederland. SUN/P.J Meertens-Instituut, Nijmegen, the Netherlands.

Karlsson, R., 2001. Svinn i storhushåll: mängder, sammansättning och möjligheter att minska. Master’s Thesis, Department of Systems Ecology, Stockholm University.

Kingmans, R., 1998. Krachtvoer wordt paar dubbeltjes duurder. Boerderij 23, 66-67.

KNMI and CHO, 1988. Van Penman naar Makkink. Een nieuwe berekeningswijze voor de klimatologische verdampingsgetallen. Commissie voor hydrologisch onderzoek TNO, ‘s-Gravenhage, the Netherlands.

Kok, R., Biesiot, W., Wilting, H.C., 1993. Energie-intensiteiten van voedingsmiddelen. Center for Energy and Environmental Studies (IVEM). IVEM-onderzoeksrapport 59. Groningen, the Netherlands.

Kok, R., Benders, R.M.J., Moll, H.C., 2001. Energie-intensiteiten van de Nederlandse consumptieve bestedingen anno 1996. Center for Energy and Environmental Studies (IVEM), IVEM-onderzoeksrapport 105. Groningen, the Netherlands.

Kok, R., Kramer, K.J., 1995. Hoeveel energie kost onze voeding? Voeding, 56(6): 12-15.

Kolk, A., 2000. Economics of Environmental Management. Financial Times/Prentence Hall.

Kramer, K.J., 2000. Food Matters. Center for Energy and Environmental Studies (IVEM). Thesis. University of Groningen, the Netherlands.

Kramer, K.J., Biesiot, W., Kok, R., Wilting, H.C., Schoot Uiterkamp, A.J.M., 1994. Energie geld(t). 1994. Interfacultaire Vakgroep Energie en Milieukunde (IVEM). IVEM-onderzoeksrapport 71. Groningen, the Netherlands.

References ___________________________________________________________________________________

100

Kramer, K.J., Moll, H.C., 1995. Energie voedt, nadere analyses van het indirecte energieverbruik van voedingsmiddelen. Center for Energy and Environmental Studies (IVEM). IVEM-onderzoeksrapport 77. Groningen, the Netherlands.

Landbouw-Economisch Instituut (LEI-DLO), Centraal Bureau voor de Statistiek (CBS), 1980.

Landbouwcijfers 1980. LEI / CBS, ‘s-Gravenhage/Voorburg/Heerlen, the Netherlands.


Landbouwcijfers 1985. ‘s-Gravenhage/Voorburg/Heerlen, the Netherlands, LEI/CBS.


Landbouwcijfers 1991. LEI / CBS, ’s-Gravenhage/Voorburg/Heerlen, the Netherlands.




Landbouwcijfers 1996. ‘s-Gravenhage/Voorburg/Heerlen, the Netherlands, LEI/CBS.




Landbouwcijfers 2002. LEI / CBS, ’s-Gravenhage.

Lang, T., 2002. Can the Challenges of Poverty, Sustainable Consumption and Good Health Governance Be Addressed in an Era of Globalization? The Nutrition Transition. Diet and Disease in the Developing World. Ed. B. Caballero and B. M. Popkin. Food Science and Technology International Series. Ed. S. L. Taylor. London: Academic Press, 51-71.

Lecoeur, J., Sinclair, T.R., 2001. Harvest index increase during seed growth of field pea. European Journal of Agronomy, 14: 173-180.

Lehner, B., Henrichs, T., Döll, P., Alcamo, J., 2001. EuroWasser. Model-based assessment of European water resources and hydrology in the face of global change. Kassel World Water Series 5, Center for Environmental Systems Research, University of Kassel, Kurt-Wolters-Strasse 3, 34109 Kassel, Germany.

Luna Bazó, P., Bracho, M., 1987. Encuesta Nacional de Nutrición. Area Socio Alimentaria ‘Encuesta de Consumo’. Mimeografiado. Instituto Nacional de Nutrición, Direccion Téchnica, Caracas, Venezuela.

Maddison, S., 2003. The world economy. Historical statistics. OECD, Paris, France.

Meadows, D.L., 1972. The limits to growth. A report for the club of Rome project on the predicament of mankind. Universe Books, New York.

References

___________________________________________________________________________________

101

Mennell, S., Murcott, A., Van Otterloo, H.A., 1992. Sociology of Food, Eating Diet and Culture. Ed. SAGE Publications. London: International Sociological Association.

Ministerio de Agricultura DANE-DRI-PAN, 1984. Encuesta Nacional de Alimentatión, Nutrición y Vivienda DANE-PAN-DRI 1981. Franza Pardo T-Bogotá (Mimeógrafo).

Ministerio de Salud, 1992. Encuesta Nacional de Consumo de Alimentos. Departamento de Nutrición y Dietética Panamá.

Ministerio de Salud, 1996. Encuesta Nacional de Nutrición. Fasciculo No 1: Consumo Aparente. San José, Costa Rica.

Moll, H.C., 1993. Energy counts and materials matter in models for sustainable development. Dynamic lifecycle modelling as a tool for design and evaluation of long-term environmental strategies. Thesis. University of Groningen, the Netherlands.

Monteith, J.L., 1977a. Climate and the efficiency of crop production in Britain. Phil. Trans. R. Soc. Lond. B., 281: 277-294.

Monteith, J.L., 1977b. Reassessment of maximum growth rates for C3 and C4 crops. Expl. Agric., 14: 1-5.

Monteith, J.L., 1994. Validity of the correlation between intercepted radiation and biomass. Agricultural and forest meteorology, 68: 213-220.

Musaiger, A.O., 1989. Changes in food consumption patterns in Bahrein. Nutrition and Health 6, 183-188.

National Institute of Nutrition (NIN), 1995. Sentinel Food and Nutrition Surveillance System data. NIN, Hanoi, Vietnam.

National Institute of Statistics (NIS/MOP), 1996. Socio-Economic Survey of Cambodja. Data from the Multi-Indicator Cluster Survey (MICS) of the Socio-Economic Survey of Cambodja (SESC) sponsored by the Asian Development Bank in collaboration with the UNICEF/UNDP/CARERE and ILO, Royal Government of Cambodja.

Netelenbos, T., Van Kemenade, J.A., Van der Ham, W.B., Tchernoff, A., 1999. Perspectievennota Verkeer en Vervoer. Ministeries van V & W, VROM, EZ, LNV, the Netherlands.

Nonhebel, S., Moll, H.C., 2001. Evaluation of Options for Reduction of Greenhouse Gas Emissions by Changes in Household Consumption Patterns. 410 200 059, 94. RIVM, National Institute of Public Health and the Environment, Bilthoven, the Netherlands.

Nonhebel, S., 2004. On resource use in food production systems: the value of livestock as ‘rest-stream upgrading system’. Ecological Economics 48: 221-30.

Oldeman, L.R., Hakkeling, R.T.A., Sombroek, W.G., 1991. World map of the status of human-induced soil degradation: global assessment of soil degradation. International Soil Reference and Information Centre, Wageningen, the Netherlands.

Organisation for Economic Co-operation and Development (OECD), 2001. Corporate Responsibility, Private Initiatives and Public Goals. OECD Publications. Paris, France.

Pallot, J., Nefedova, T., 2003. Trajectories in people’s farming in Moscow oblast during the post- socialist transformation. Journal of rural studies 19, 345-362.

References ___________________________________________________________________________________

102

Penning de Vries, F.W.T., 1983. Modeling of growth and production. [12D]. Encyclopedia of Plant Physiology, New Series. Springer Verlag, Berlin.

Penning de Vries, F. W. T., Jansen, D.M.; Ten Berge, H.F.M.; Bakema,A.l., 1989. Simulation of Ecophysiological Processes of Growth in Several Annual Crops. Wageningen, the Netherlands: Centre for Agricultural Publishing and Documentation (Pudoc), 63-64.

Penning de Vries, F.W.T., Van Keulen, H., Rabbinge, R., 1995. Natural resources and limits of food production in 2040. In: J.Bouma (Editor), Eco-regional approaches for sustainable land use and food production. Kluwer Academic Publishers, the Netherlands.

Pimentel, D., 1980. Handbook of energy utilization in agriculture. CRC Press, Inc., Boca Raton, Florida, USA.

Pimentel, D., Houser, J., 1997. Water resources: agriculture, the environment, and society. Bioscience, 47(2): 97-107.

Pindyck, R.S., Rubinfeld, D.L., 2005. Microeconomics, sixth edition. Pearson Education, Inc., Upper Saddle River, New Yersey, 07458, USA.

Pinstrup-Andersen, P., 2000. Food policy research for developing countries: emerging issues and unfinished business. Food Policy 25, 125-141.

Popkin, B. M. The Dynamics of the Dietary Transition in the Developing World. The Nutrition Transition. Diet and Disease in the Developing World. Ed. B. Caballero and B. M. Popkin. Food Science and Technology International Series. Ed. S. L. Taylor. London: Academic Press, 2002. 111-29.

Produktschap Margarine, Vetten en Oliën, 1993. Statistisch jaarboek 1993. Rijswijk, the Netherlands.

Rabbinge, R., Van Latesteijn, H.C., 1992. Long term options for land use in the European community. Agricultural Systems, 40: 195-210.

Receveur, O., Boulay, M., Kuhnlein, H.V., 1997. Decreasing Traditional Food Use Affects Diet Quality for Adult Dene/Métis in 16 Communities of the Canadian Northwest Territories. Journal of Nutrition, 127: 2179-2186.

Renault, D., Wallender, W.W., 1999. Nutritional water productivity and diets. Agricultural water management, 45: 275-296.

Rockstrom, J., 1999. On-Farm Green Water Estimates as a Tool for Increased Food Production in Water Scarce Regions. Phys. Chem. Earth (B), 24(4): 375-383.

Rondinelli, D., Vastag, G., 2000. Panacea, Common Sense, or Just a Label? The Value of ISO 14001 Environmental Management Systems. European Management Journal, 18(5): 499-510.

Rondinelli, D.A., Berry, M.A., 2000. Environmental Citizenship in Multinational Corporations: Social Responsibility and Sustainable Development. European Management Journal, 18(1): 70-84.

Rosegrant, M.W., Ringler, C., 1998. Impact of food security and rural development of transferring water out of agriculture. Water Policy, 1: 567-586.

References

___________________________________________________________________________________

103

Schmidt-Nielsen, K., 1988. Animal Physiology: Adaptation and Environment. third ed. New York: Press Syndicate of the University of Cambridge.

Schulze, R.E., 2000. Modelling hydrological responses to land use and climate change: a Southern African perspective. Ambio, 29(1): 12-22.

SER, 2001. Corporate social responsibility. Assen, the Netherlands, Royal van Gorcum.

Sibma, L., 1968. Growth of closed green crop surfaces in the Netherlands. Neth. J. of Agric. Sc., 16: 211-216.

Simeon, D.T., Patterson, A.W., CFNI, 1994. Energy and protein accessibility at the household level in Jamaica: results from a national survey 1989.

Smil, V., 2002. Worldwide transformation of diets, burdens of meat production and opportunities for novel food proteins. Enzyme and Microbial Technology 30, 305-11.

Smith, M., Allen, R., Monteith, J.L., Pereira, L.A., Perrier, A., Segeren, A., 1991. Report on the Expert Consultation for the Revision of FAO methodologies for crop water requirements. FAO/AGL, Rome.

Spedding, C.R.W., 1988. An introduction to agricultural systems. Elsevier Applied Science, Amsterdam, the Netherlands.

Stauffer, J.E., 1997. ISO 14000 Standards. In: Cereals-Foods-World. pp. 228-230.

Steg, L., Vlek, C., Feenstra, D., Gerbens-Leenes, P.W., Karsten, L., Kok, R., Lindenberg, S., Maignan, I., Moll, H., Nonhebel, S., Schoot Uiterkamp, A.J.M., Sijtsma, F., and Witteloostuijn, A., 2001. Towards a comprehensive model of sustainable corporate performance, three-dimensional modelling and practical measurement. University of Groningen, departments of economics, environmental sciences, management science, psychology and sociology. Groningen, the Netherlands.

Statistisches Bundesamt, 1998. Entwicklung von physischen Wasserflussberechnungen für Deutschland. Endbericht. Im Auftrag von Eurostat.

Steger, U., 1993. The greening of the board room: how German companies are dealing with environmental issues. In: K. Fischer and J. Schot (Eds.), Environmental Strategies for Industry. Island Press, Washington, DC.

Thoresen, J., 1999. Environmental performance evaluation - a tool for industrial improvement. Journal of Cleaner Production, 7: 365-370.

Tilman, D., Cassman, K.G., Matson, P.A., Naylor, R., Polasky, S., 2002. Agricultural sustainability and intensive production practices. Nature, 418: 671-677.

Tu Giay, Chu Quoc Lap, 1990. Final report on the subject 64D.01.01 of the National Research Programme Natioanl General Survey 1989. The Governmental Science and Technology Committee, NIN, Hanoi, Vietnam.

United Nations, 1998. World Populations Projections to 2150. Dept. of Int. Economic and Social Affairs, New York.

United Nations Population Division, 2002. World Population Prospects: The 2002 Revision. http://www.un.org/esa/population/publications/wpp2002/wpp2002annextablesPDF (13-10-2003).

References ___________________________________________________________________________________

104

United Nations, 2005. The Millennium Development Goals Report, 2005. United Nations, New York.

United Nations Statistics Division, 2006. http://unstats.un.org/unsd/methods/icp/gdp/gdp06_htm.htm

United States Department of Agriculture (USDA) Agricultural Research Service, 2005. Food and Nutrient Database for Dietary Studies, 1.0. http://www.ars.usda.gov/Services/docs.htm?docid=7637 30th November 2005.

Veleva, V., Ellenbecker, M., 2001. Indicators of sustainable production: framework and methodology. Journal of Cleaner Production, 9: 519-549.

Vereniging van Nederlandse koffiebranders en theepakkers, 1998. Jaarverslag 1998. Amsterdam, the Netherlands.

Verkerk, G., Broens, J.B., Kranendonk, W., Puijl van der, F.J., Sikkema, J.L., Stam, C.W., 1986. Binas, informatieboek vwo-havo voor het onderwijs in de natuurwetenschappen. tweede druk., Wolters-Noordhof bv., Groningen, the Netherlands.

Voedingscentrum, 1998a. Nederlandse Voedingsmiddelentabel. Breedveld, B. C., Hammink, J., Oosten van, H. M. (Eds.). Den Haag, the Netherlands.

Voedingscentrum, TNO, 1998b. Zo eet Nederland 1998. Voedingscentrum, Den Haag, the Netherlands.

Voorlichtingsbureau voor de Voeding, 1973. Nederlandse voedingsmiddelentabel. ’s Gravenhage, the Netherlands.



Vollenbroek, F.A., 2002. Sustainable development and the challenge of innovation. Journal of Cleaner Production, 10: 215-223.

Vringer, K., Blok, K., 1995. The direct and indirect energy requirement of households in the Netherlands. Energy Policy 23 (10), 893-910.

Vuuren van, D.P., Smeets, E.M.W., Kruijf de, H.A.M., 1999. The ecological footprint of Benin, Bhutan, Costa Rica and the Netherlands. Rijksinstituut voor Volksgezondheid en Milieuhygiene (RIVM), Bilthoven, the Netherlands.

Vuuren van, D.P., Smeets, E.M.W., 2000. Ecological footprints of Benin, Bhutan, Costa Rica and the Netherlands. Ecological Economics, 34(234): 115-130.

Wackernagel, M., Onisto, L., Callejas Linares, A., Lopez Falfan, I.S., Mendez Garia, J., Suarez Guerrero, A.I., Suarez Guerrero, M.G., 1997. Ecological Footprints of Nations: How Much Nature Do They Use? How Much Nature Do They Have? International Council for Local Environmental Initiatives, Toronto, Canada.

Wallace, J.S., 2000. Increasing agricultural water use efficiency to meet future food production. Agriculture, Ecosystems & Environment, 82: 105-119.

References

___________________________________________________________________________________

105

Wandel, M., 1988. Household food consumption and seasonal variations in food availability in Sri Lanka. Ecology of Food and Nutrition, 22: 169-182.

Wegener Sleeswijk, A., Meeusen-van Onna, M.J.G., Sengers, H.H.W.J.M., Van Zeijts, H., Reus, J.A.W.A., 1996. Toepassing van LCA voor agrarische produkten. Den Haag, the Netherlands, Centrum voor Milieukunde Leiden, Centrum voor Landbouw en Milieu, Landbouw-Economisch Instituut (LEI-DLO).

Weidema, B.P., 1999. Data, databases and software for LCAs on food. Theme report for the LCA-NET-Food.

Weidema, B.P., Meeusen, M.J.G., 2000. Agricultural data for Life Cycle Assessments. 2.00.01. The Hague, the Netherlands, Agricultural Economics Research Institute (LEI).

Whitman, W.E., Holdsworth, S.D., 1975. Water use in the food industry. Campden Food Preservation Research Association.

Whitney, E.N., Rolfes, S.R., 1999. Understanding nutrition. Wadsworth publishing Company, Belmont, USA, pp.3-4, 40.

Williams, H., Medhurst, J., Drew, K., 1993. Corporate strategies for a sustainable future. In: K.Fischer and J.Schot (Eds.), Environmental Strategies for Industry. Island Press, Washington, DC.

Wilting, H.C., 1996. An energy perspective on economic activities. Center for Energy and Environmental Studies, University of Groningen, the Netherlands.

Wilting, H.C., Benders, R.M.J., Biesiot, W., Louwes, M., Moll, H.C., 1999. Energy analysis program, manual version 3.0. Groningen, the Netherlands, Center for Energy and Environmental Studies (IVEM). IVEM-onderzoeksrapport, 98.

Woodward, J., 1699. Some thoughts and experiments concerning vegetation. In: Philosophical Transactions of the Royal Society.

World Bank, 2005a. East Asia's Dollar Influx-Signal for Change. Special focus. Gender Equality in East Asia.

World Bank, 2005b. https://publications.worldbank.org/subscriptions/WDI

World Business Council for Sustainable Development, 1997. Signals of Change: Business Progress Toward sustainable Development. Geneva, Switzerland, WBCSD.

World Commission on Environment and Development, 1987. Our Common Future. Oxford, UK, Oxford University Press.

Wijn de, J.F., 1968. De veranderingen in het Nederlandse voedingspatroon. Voeding, 29(11): 490-504.

Wijn de, J.F., Weits, J., 1971. Steensma’s Voedingsleer. Scheltema en Holkema, Amsterdam / Haarlem, the Netherlands.

Wit de, C.T. Transpiration and Crop Yields. 64.6. 1958. Pudoc. Agricultural Research Reports. Wageningen, the Netherlands.

Wit de, C.T., 1992. Resource use efficiency in agriculture. Agricultural Systems, 40: 125-151.

References ___________________________________________________________________________________

106

Zhang, H., Oweis, T.Y., Garabet, S., Pala., 1998. Water-use efficiency and transpiration efficiency of wheat under rain-fed conditions and supplemental irrigation in a Mediterranean-type environment. Plant and Soil 201, 295-305.

107

GGlloossssaarryy CVF Conversion factor

EMAS European Union's Eco-Management and Audit Scheme

EMS Environmental Management Systems

ERE Energy Required for Energy

EU European Union

FAO Food and Agricultural Organisation of the United Nations

GRI Global Reporting Initiative

H-crops Hypothetical crops

HEI High Environmental Impact

IPCC International Panel on Climate Change

ISO International Standards Organisation

LCA Life Cycle Assessment

LCSP Lowell Center for Sustainable Production

MJ Megajoule

MT Metric tons (1 metric ton = 1000 kg)

OECD Organisation for Economic Co-operation and Development

RUE Radiation use efficiency

SCP Sustainable Corporate Performance

108

109

AAppppeennddiixx AA

LLand requirements, indirect energy requirements and household requirements for food items in the Netherlands in 1990. Table A1 Land and indirect energy requirement per food item (m2 and MJ per kg) available in the Netherlands in 1990, and Dutch household requirements in 1990 (kg per household per year)

Food item Land requirement (m2

per kg)

Energy requirement

(MJ per kg) [1]

Household requirement (kg per household per

year) [2, 3]

Meat and fish

Meat

Deep frozen meat 15.7 109.8 3.9 Roast meat loaf (package)

15.7 96.0 0.6

Pork (canned) 6.0 49.6 0.8 Beef (canned) 9.7 58.6 0.8 Poultry (canned) 5.1 46.5 0.8 Chicken filet (fresh) 10.2 86.0 10.2 Ham (package) 12.0 102.7 3.9 Bacon (package) 12.0 93.5 1.7 Other sausages and meat products (package)

14.4 96.3 0.5

Horse meat (fresh) 19.4 125.4 0.5 Smoked beef (package)

30.9 141.9 0.3

Smoked horse meat (package)

30.9 119.4 0.3

Bacon (lean, package)

12.0 82.8 1.7

Minced meat (fresh, beef, package)

19.4 104.9 8.0

Minced meat (fresh, beef/pork, package)

15.7 91.9 8.0

Veal (fresh, steak) 1.3 189.5 0.5 Beef (fresh, streaky) 19.4 110.4 8.4 Pork (fresh, chops) 12.0 86.4 15.1 Game (e.g. rabbit, fresh)

0.0 35.6 0.2

Sausages (fresh) 14.4 91.0 14.9

Fish Fresh fish (e.g. cod) 0.0 79.4 2.4 Deep frozen fish (pollack, cardboard box)

0.0 133.0 1.4

Fried fish 0.0 115.5 0.7 Smoked herring 0.0 124.4 0.3 Fish (e.g. salmon, canned)

0.0 105.8 1.1

Salt herring 0.0 53.3 1.4

Appendix A ___________________________________________________________________________________

110

Pickled herring (glass jar)

0.0 49.7 0.4

Dairy and eggs Eggs (cardboard box)

4.9 24.3 13.4

Cheese (48+, new) 6.8 71.5 27.6 Cheese (48+, ripe) 7.0 79.5 -- Cheese (48+, old) 7.2 89.7 -- Cheese (brie) 7.3 83.1 -- Cheese (48+, cumin- seed)

6.8 74.4 --

Buttermilk (cardboard package)

0.7 10.8 25.3

Condensed milk (semi-skimmed, liter, glass bottle)

2.2 19.5 9.8

Condensed milk (semi-skimmed, liter, cardboard package)

2.2 20.1 --

Condensed milk (full fat, liter, glass bottle)

2.8 22.1 2.0

Condensed milk (full fat, cardboard package)

2.8 22.7 --

Milk (semi-skimmed, cardboard package)

0.9 8.0 116.1

Milk (semi-skimmed, liter, glass bottle)

0.9 7.7 --

Milk (skimmed, liter, cardboard package)

0.7 8.4 5.1

Milk (full fat, liter, cardboard package)

1.3 10.2 34.6

Milk (full fat, liter, glass bottle)

1.3 9.9 --

Chocolate custard (full fat, liter, cardboard package)

1.5 11.5 6.7

Chocolate custard (full fat, liter, glass bottle)

1.5 10.9 --

Cream (cardboard package)

6.7 39.9 2.7

Cream (glass bottle) 6.7 39.0 -- Dairy butter (aluminum wrapper)

14.9 82.6 3.3

Dairy butter (paper wrapper)

14.9 82.0 --

Fruit yogurt (full fat, liter, cardboard package)

1.3 13.6 18.3

Fruit yogurt (full fat, liter, glass bottle)

1.3 13.0 --

Appendix A

___________________________________________________________________________________

111

Fruit yogurt (skimmed, liter, cardboard package)

0.7 11.6 --

Fruit yogurt (skimmed, liter, glass bottle)

0.7 11.0 --

Yogurt (skimmed, liter, cardboard package)

0.7 7.8 29.8

Yogurt (skimmed, liter, glass bottle)

0.7 7.0 --

Yogurt (full fat, liter, cardboard package)

1.3 10.7 10.1

Yogurt (full fat, liter, glass bottle)

1.3 10.0 --

Oils and fats

Cooking and frying fat

8.9 14.4 8.0

Low fat spread (tub) 3.8 16.7 9.8 Margarine (tub) 8.9 18.1 20.6 Margarine (wrapper) 8.9 14.4 -- Vegetal oil 10.9 28.7 4.4 Beverages Beer (lager, can 0.33 l)

0.5 9.3 77.3

Beer (lager, bottle 0.33 l)

0.5 8.1 --

Cocoa 3.0 24.0 0.3 Strong alcoholic drinks, 30%

0.02 27.7 6.9

Coffee 15.8 38.7 8.9 Mineral water (without package)

0.0 3.3 25.9

Soft drinks (glass bottle, 1 l)

0.3 6.9 110.7

Soft drinks (PE bottle, 1.5 l)

0.3 6.5 --

Soft drinks (PE bottle, 2 l)

0.3 6.4 --

Soft drinks (glass bottle, 0.2 l)

0.3 7.8 --

Soft drinks (can, 0.33 l)

0.3 11.5 --

Tea (bags in cardboard box)

35.2 51.1 1.8

Orange juice (glass bottle)

0.9 15.9 56.4

Orange juice (cardboard package)

0.9 13.9 --

Red wine 1.5 33.2 29.3

Appendix A ___________________________________________________________________________________

112

White wine 1.5 33.2 -- Bread Rusk 2.3 42.9 4.4 Buns 1.3 30.8 9.9 Brown bread 1.1 12.4 73.6 Raisin buns (plastic bag)

1.3 46.7 2.8

Fruit loaves and rolls

1.3 30.7 8.5

White bread 1.3 13.1 47.9 Potatoes, vegetables and fruits

Potatoes Potatoes 0.2 1.8 153.6 Potatoes (plastic bag)

0.2 1.9 --

French fries (deep frozen)

0.4 52.8 --

Vegetables Chicory and lettuce (greenhouse)

0.04 46.6 11.6

Chicory and lettuce (open air)

0.2 7.6 --

Chicory and lettuce (organic)

8.1 --

Spinach (greenhouse)

0.1 47.8 2.8

Spinach (open air) 0.3 7.6 -- Spinach (organic) 8.1 -- Other leaf vegetables (open air)

0.2 7.6 12.2

Other leaf vegetables (open air, package)

0.2 9.1 --

Other leaf vegetables (organic)

8.1 --

Cauliflower (greenhouse)

0.1 46.6 11.1

Cauliflower (open air)

0.5 7.6 --

Cauliflower (organic) 8.1 -- Mushrooms 0.1 18.4 3.1 Vegetables (can) 0.1 12.4 24.0 Vegetables (glass, deposit)

0.1 12.8 --

Vegetables (glass, disposable)

0.1 14.0 --

Vegetables (deep frozen, cardboard

0.4 25.6 6.2

Appendix A

___________________________________________________________________________________

113

package) Vegetables (deep frozen, plastic package)

0.4 23.0 --

Vegetables (dried) 9.0 43.0 0.6 Cucumber (greenhouse)

0.02 38.8 3.8

Cucumber (open air) 0.2 7.6 -- Cucumber (organic) 8.1 -- Red cabbage (open air)

0.1 7.6 17.4

Red cabbage (organic)

8.1 --

Peppers (greenhouse)

0.1 75.0 5.8

Pulses (beans) 3.5 7.4 -- Pulses (peas) 2.0 7.4 -- Pulses (lentils) 6.3 7.4 -- Pulses (soy beans) 5.2 7.6 -- French beans (greenhouse)

0.3 46.6 6.0

French beans (open air)

0.9 7.6 --

French beans (organic)

8.1 --

Sprouts (open air) 0.5 7.6 2.3 Sprouts (organic) 8.1 -- Tomatoes (greenhouse)

0.03 50.2 4.9

Tomatoes (open air) 0.2 7.6 -- Tomatoes (organic) 8.1 -- Onions 0.2 8.1 10.5 Chicory 0.7 12.0 -- Carrots (open air) 0.2 7.6 3.6 Carrots (open air, package)

0.2 9.1 --

Carrots (organic) 8.1 -- Celeriac (open air) 0.3 7.6 -- Celeriac (open air, packed)

0.3 9.1 --

Celeriac (organic) 8.1 -- Sauerkraut (tub) 0.2 14.2 3.0 Sauerkraut (plastic package)

0.2 14.6 --

-- Fruits Strawberries (greenhouse, cardboard box)

0.5 56.1 3.2

Strawberries (open air, cardboard box)

0.6 11.4 --

Strawberries (organic, cardboard box)

11.8 --

Apples (plastic bag) 0.4 10.4 39.4 Apples (cardboard 0.4 11.8 --

Appendix A ___________________________________________________________________________________

114

box) Apples (plastic bowl) 0.4 10.8 -- Bananas 0.2 14.4 15.4 Red currants 1.5 13.5 0.4 Grapefruits 0.4 15.8 3.9 Grapes 1.0 15.2 3.4 Jam 1.1 27.8 5.9 Cherries 4.2 13.8 1.5 Mandarins 0.5 15.8 7.6 Melons (greenhouse)

0.2 54.4 3.3

Melons (open air) 0.8 9.8 -- Pears 0.6 9.8 5.8 Peaches 0.7 15.4 3.4 Prunes 0.9 14.7 1.1 Oranges 0.5 15.8 33.7 Apple sauce (can) 0.5 21.2 14.6 Apple sauce (glass, disposable)

0.5 22.7 --

Apple sauce (glass, deposit)

0.5 18.7 --

Raisins 4.3 79.5 1.5 Fruits in juice (can) 0.5 25.6 4.7 Fruits in juice (glass, disposable)

0.5 27.1 --

Cakes, pastries and chips

Chips 0.9 49.6 4.2 Cakes and pastries 1.4 29.0 61.1

Pasta and flour products

Potato starch (cardboard box)

1.6 15.8 1.2

Pasta (plastic bag) 1.9 13.5 7.1 Pasta (cardboard box)

1.9 14.5 --

Groceries (paper bag)

1.6 17.7 9.9

Groceries (cardboard box)

1.6 19.2 --

Rice (cardboard box)

3.4 21.9 6.5

Rice (plastic bag) 3.4 20.9 -- Wheat flour (paper bag)

1.9 12.5 9.4

Wheat flour (paper bag)

1.6 11.0 --

Other food items

Appendix A

___________________________________________________________________________________

115

Chocolate bar 2.0 58.4 5.5 Chocolate vermicelli’s (cardboard box)

2.0 28.2 4.3

Chocolate vermicelli’s (plastic bag)

2.0 26.5 --

Honey (glass pot, metal lid)

1.2 32.0 1.0

Honey (glass pot, plastic lid)

1.2 33.9 --

Main coarse dishes (can) [4]

2.6 90.8 4.5

Main coarse dishes (deep frozen) [4]

2.6 95.8 --

Main coarse dishes (cooled, plastic tray) [4]

2.6 57.8 --

Nuts and peanuts (cardboard box)

0.8 42.7 7.7

Nuts and peanuts (plastic bag)

0.8 43.1 --

Peanut butter (glass pot)

0.6 36.8 2.1

Sauces and relish (25% oil, glass pot)

1.8 35.3 11.0

Sauces and relish (25% oil, plastic bottle)

1.8 37.8 --

Sauces and relish (40% oil, tube)

1.8 49.7 --

Soup (main coarse dish, can) [5]

1.9 21.7 13.9

Syrup (glass pot, disposable)

0.9 18.2 0.3

Syrup (cardboard pot)

0.9 21.2 --

Sugar (paper box) 1.2 19.1 22.3 Sugar (paper bag) 1.2 17.2 -- Vegetarian meat substitutes (package)

2.6 56.4 --

Sugar products on bread (cardboard box)

1.2 30.9 11.5

Sugar products on bread (plastic bag)

1.2 29.2 --

[1] Source: Kramer and Moll, 1995 [2] In 1990, a Dutch household consisted of 2.41 persons [3] When food items are available in more than one type of package or are produced in more than one production system, household requirements for that food type are shown including all package and production types [4] Ingredients: 183 g potatoes, 183 g boiled rice, 183 g boiled pastry; 93 g beef, 93 g pork, 15 g margarine and 110 g vegetables for 1 kilo product

Appendix A ___________________________________________________________________________________

116

[5] Ingredients: 40 g pastry, 10 g vegetables, 20 g sugar, 102 g pork and 102 g beef for 1 kg soup -- No data available

117

AAppppeennddiixx BB

OOverview of the fifty two countries for which Chapter 6 performed a spatial analysis

Africa

Algeria, the Democratic Republic of Congo, Ivoiry Coast, Egypt, Ethiopia, Kenya, Morocco,

Nigeria, Sudan, Tanzania, and South Africa

Asia

Bangladesh, China, India, Indonesia, Malaysia, Pakistan, the Philippines, Sri Lanka, Thailand,

and Vietnam

Latin America

Argentina, Brazil, Chile, Colombia, Ecuador, Guatemala, Mexico, Peru, and Venezuela

Middle East

Israel

OECD

Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan,

the Netherlands, Portugal, Spain, Sweden, Turkey, United Kingdom, United States

Additional, small countries

The United Arab Emirates, Estonia, Slovenia.

118

119

AAppppeennddiixx CC

OOverview of countries with national food surveys used in Chapter 6 and their authors.

Nineteen food surveys that provide data on fat E% are marked with an asterix *.

Argentina

Britos, S., Scacchia, S., 1998.

Bangladesh

Jahan and Hossein, 1998*.

Brazil

Galeazzi, M.A.M., Falchoni, P. Jnr., 1998.

Cambodja

National Institute of Statistics (NIS/MOP), 1996.

China

Institute of Nutrition and Food Hygiene (INFH), 1985.

Ge, K., Zhai, F., Yan, H., 1996*.

.

Colombia

Ministerio de Agricultura DANE-DRI-PAN, 1984.

Costa Rica

Ministerio de Salud, 1996.

Egypt

Hassanyn, A.S., 2000*.

El Salvador

Asociación Demográfica Salvadorĕna (ADS), Ministerio de Salud Pública y Asistencia Social

(MSPAS), Instituto de Nutrición de Centro América y Panamá (INCAP), 1990.

Equador

Freire, W., 1988.

Guinee

FAO, 2004b*.

Appendix C ___________________________________________________________________________________

120

Iran

Djazayery, A., Samimi, B., 1996. (Surveys for 1983* and 1992*)

Jamaica

Simeon, D.T., Patterson, A.W., CFNI, 1994.

Jordan

Department of Statistics (DOS), 1997*.

Madagaskar

FAO, 2004b

Mali

FAO, 2005*.

Mexico

INNSZ, 1990*.

Avila, A., Shamah, T., Chavez, A., 1997*.

The Netherlands

Voedingscentrum, TNO, 1998*.

Panama

Ministerio de Salud, 1992*.

Instituto de Nutrición de Centro América y Panamá (INCAP), Oficina de Investigaciones

Internationales de Salud, Ministerio de Salud Pública y Asistencia Social (MSPAS), 2000.

Peru

Amat, C., Curonisy, P., 1981.

Philippines

Food and Nutrition Research Institute of the Department of Science and Technology (FNRI-

DOST) of the Philippines, 2000*.

Sri Lanka

Department of Census and Statistics, 1993*

Turkey

Hundd and Moh, 1997*.

Appendix C ___________________________________________________________________________________

121

United States

United States Department of Agriculture (USDA) Agricultural Research Service, 2005*

Venezuela

Luna Bazó, P., Bracho, M., 1987*.

Vietnam

Tu Giay, Chu Quoc Lap, 1990*

National Institute of Nutrition (NIN), 1995*.

Zimbabwe

Bursztijn, P.G., 1985.

122

123

SSuummmmaarryy

11 ►► IInnttrroodduuccttiioonn Humans need food to remain healthy and stay alive. Worldwide, the land-based manufacture of food puts a large claim on scarce natural resources, such as land, fresh water, and energy carriers. This thesis emphasizes the relationship between food consumption patterns and the use of natural resources. The three objectives are (i) to design and develop measuring methods for environmental sustainability in food production systems, (ii) to analyze the effect of food consumption patterns on natural resources, and (iii) to identify those food consumption patterns that have the most sustainable environmental performance under nutritional and cultural constraints. The thesis continues one of the lines of research of the Center for Energy and Environmental Studies (IVEM) and examines sustainability from the perspective of consumers. It provides directions of change to make food consumption patterns more sustainable, in developed as well as in developing countries. 22 ►► FFoooodd ssyysstteemmss A food system consists of a consumption and a production system. The size of a population, in combination with the per capita types and amounts of specific foods consumed, i.e. food consumption patterns, determines the total amount of food commodities and food items needed in a country. Food consumption patterns differ strongly among countries and change in time. Globally, the main difference occurs between food consumption patterns in developing and in developed countries. Developing countries show patterns that are mainly based on staples, whereas the patterns in developed countries have more variation and contain larger amounts of affluent foods, such as meat. In some developing countries, for example in China and India, consistent economic growth occurs, growth that is accompanied by shifts in food consumption towards the affluent patterns of developed countries. Due to the large populations of these countries, such a shift might have huge consequences for the impact on natural resources. Production systems include several production chains made up of links. Agriculture, the first link of a production chain, needs land and fresh water to produce food. In developed countries after World War ll, large yield increases have occurred due to larger application of fertilizers, pesticides, irrigation, and the introduction of new crop varieties. As a result, land requirements per unit of output have decreased. However, increased output per unit of land requires more (fossil) energy input. Beside land and energy, fresh water is also a prerequisite for crop growth and for livestock. Today, agriculture already needs 70 percent of human fresh water requirements, mainly for irrigation. However, increasingly, agricultural water use has to compete with urban and industrial water use. While land and water are mainly needed in the first link of a food production chain, agriculture, energy is needed in all links of a chain: agriculture, the food industry, trade, retailing, as well as for transportation between chain links. The requirement of a food item for natural resources depends on the characteristics of a food production system. This means that the total requirement of a food consumption pattern for natural resources depends on a unique combination of a specific production and consumption system. 33 ►► MMeeaassuurriinngg mmeetthhoodd ffoorr eennvviirroonnmmeennttaall ssuussttaaiinnaabbiilliittyy iinn ffoooodd pprroodduuccttiioonn ssyysstteemmss Corporate responsibility and sustainable business practices are important issues for private companies. Society demands that companies not only pay attention to economic interests but also to collective long term interests concerning human well being and the environment. Sustainable business practices have three dimensions, the economic, the social, and the environmental dimension. This thesis mainly addresses the environmental dimension. Environmental impacts of company activities take place on several levels of scale, from the local to the global level. The influence of an individual company on the local environment can be large, but decreases when environmental impacts take place on higher levels of scale, for example, the impact of carbon dioxide emissions on climate change. This means that there is a need for shared responsibility when more companies contribute to a final product, and impacts take place on the global scale level.

Summary ___________________________________________________________________________________

124

At present, it is difficult to assess sustainable business practices because there is no internationally accepted standard on what, when, and where to report. Moreover, companies are often part of complex production systems. This sometimes means that large efforts in a specific company to decrease environmental impacts may still result in small improvements in a production system as a whole, or even in larger impacts in other companies. The design and development of a measuring method for environmentally sustainable business practices, therefore, requires a system approach that includes all companies that contribute to a final product. For food, research and companies have performed many efforts to measure environmentally sustainable production, but have often focused on events on a local level. This has generated a large number of indicators and data sets that often do not provide insight into the environmental sustainability of the food system and ignores interactions among indicators. This thesis proposes the use of only three indicators that address global issues: the use of land, fresh water and energy (from both fossil and renewable sources). The systemic approach can show trade-offs along supply chains that make up a production system. The use of the method implies an extension of environmental performance of a company towards the overall performance of a production system. The final outcome shows the total land, energy, and water requirement per kilogram of available food. Companies can use the data to compare trends over time and results with targets, and to benchmark a company against others. Consumers can use the data to compare environmental effects of various foods. The thesis has designed the method for food production systems but it is also applicable for other business sectors. Acceptance of this measuring method may be a powerful contribution towards creating sustainable business practices. 44 ►► MMeetthhoodd ttoo ddeetteerrmmiinnee llaanndd rreeqquuiirreemmeennttss ffoorr ffoooodd ccoonnssuummppttiioonn ppaatttteerrnnss Food production requires cropland, land for horticulture, and grazing land for cattle. Globally, the per capita amount of land available for food production is declining. Between 1961 and 1998, for example, the per capita availability of land for food decreased from 1.5 to 0.8 hectares. Over the last decade, several studies were published on the issue of food security. Available studies have shown that future generations can be fed but the studies have simplified consumption. They have estimated that an affluent diet requires more than three times as much land as a vegetarian diet. Other differences in food consumption patterns, though, also have effects on land requirements. Increasing affluence in society causes a shift in consumption patterns, including food. Insight into the effect of these shifts on land requirements requires detailed information on land requirements of food items. This thesis assesses the impact of consumption patterns on land requirements for food. First, it develops a method to calculate land required to produce individual food items. In combination with data on consumption, it determines household land requirements for food. Applied for the Dutch situation in 1990, the method presents land requirements for over a hundred food items (see also the dataset in Appendix A). Especially foods associated with affluent food consumption patterns, such as meat, dairy butter, vegetal fats, coffee, and tea have large land requirements (m2 per kg), while food items that derive from arable farming and horticulture, such as bread, beer, sugar, potatoes, and vegetables, have relatively small land requirements. Beside the land requirement per unit of food, total consumption of the food determines its land requirement. Food items that have large land requirements per unit of food, but small consumption, for example tea, do not contribute substantially to the total land requirement of a consumption pattern. On the other hand, food items that have relatively small land requirements per unit of food, but large consumption, for example beer, contribute substantially. The combination of data on land requirements for food items and data on consumption provides information on land requirements for food consumption patterns. Results show that in 1990 a Dutch household needed the largest fraction of the land area for meat (29 percent), vegetal oils and fats (24 percent), dairy and eggs (17 percent) and for coffee, tea, wine, and beer (11 percent). Only a small fraction was needed for potatoes, vegetables and fruits (5 percent), and for other foods, such as bread (9 percent). The land area required to provide a humans food need, however, depends on a specific combination of a production and a consumption system. It is stressed that results are typical for the

Summary ___________________________________________________________________________________

125

Netherlands in 1990, and can therefore not be used to derive land requirements for other populations. However, the method can be applied to other countries for which required data are available. In this way, the thesis provides a tool for the evaluation of other combinations, and, in this way, contributes to the discussion on future land use. 55 ►► CCoonnssuummppttiioonn ppaatttteerrnnss aanndd llaanndd rreeqquuiirreedd ffoorr ffoooodd

Consumption patterns have large impacts on land required for food. The decreasing per capita availability of land requires a better insight into the relation between consumption and land requirements. The thesis combines data on land requirements per food item for the Dutch 1990 production system with data on per capita food consumption of various food consumption patterns, varying from subsistence to affluent, leading to information on differences among relative land requirements for food consumption patterns. Results of the comparison, therefore, do not provide actual square meters committed to the production of food, but relative land requirements that show differences in the impact of consumption on land requirements. The thesis showed that there are large differences among per capita food consumption patterns. In the Netherlands over the period 1950-1990, for example, average consumption showed large changes. In 1950, consumption patterns were characterized by relatively small consumption of expensive food items, such as meat, fruits, alcoholic beverages, coffee, and cheese, and relatively large consumption of cheap food items, such as potatoes. In 1990, consumption of cheaper food items had shifted towards consumption of more expensive ones. When data of the Dutch production system were used as input, this change of consumption had resulted in an increase of the relative land requirement of 40 percent. In 1995, the countries of the European Union showed differences in per capita food consumption patterns, resulting in differences in relative land requirements. Based on Dutch production data of 1990, the average pattern in Portugal had the smallest relative land requirement, the average pattern in Denmark the largest, 40 percent more. When differences per consumption category are taken into account, differences were even larger. The sum of the larges values of the five consumption categories (beverages; fats; meat; dairy and eggs; and cereals, flour, sugar, potatoes, vegetables, and fruits) differed by a factor of two from the sum of the smallest values, while the land requirement for a hypothetical diet based on wheat was six times less than that for the affluent pattern in Denmark. Especially in developing countries, people still have consumption patterns that are mainly based on staples, requiring relatively little land. However, trends towards the consumption of foods associated with affluent lifestyles will bring with them a need for more land. In the near future, changes in consumption patterns, rather than population growth, might form the most important variable for total land requirements for food. 66 ►► CCoonnssuummppttiioonn ppaatttteerrnnss aanndd wwaatteerr rreeqquuiirreedd ffoorr ffoooodd Food production requires about 70 percent of worldwide, human, fresh water use. It is possible that in some areas in the near future, the required increase of food production will lead to water shortages. Many studies have made assessments of water requirements for specific situations to provide a yield, sometimes also calculating water requirements for individual food items. The use of different system boundaries, however, has led to incompatible results. Insight into the effect of consumption patterns on water required for food, similar to land, needs information on specific water requirements for individual food items. This thesis develops a generally applicable method for the assessment of crop growth-related water flows, or ‘transpirational’ water requirements of agricultural crops. While agricultural studies have made models to assess water requirements for practical situations, this thesis uses the information the other way around. It argues that water has to be present for growth to occur. Based on the linearity of processes taking place in all green plants, the thesis develops a method to calculate the growth-related factor of crop water requirements, assesses the impact of crop characteristics on water requirements, and evaluates options to reduce the use of water by changing food consumption patterns. The thesis calculates ‘transpirational’ water requirements for a representative group of crops with different functions for human nutrition, such as staple crops, vegetables, and livestock fodder (liters per kg of fresh weight, liters per unit of dry matter, and liters per kJ of nutritional energy).

Summary ___________________________________________________________________________________

126

Because of the large difference in water content among crops, ‘transpirational’ water requirements per unit of fresh weight show large variations. Water requirements per unit of dry mass, however, differ only by a factor of two, a difference that arises from differences among harvest indices of crops and their chemical composition. Differences in requirements per unit of nutritional energy are small. This means that there are few options to reduce the use of water by substitution of crops in food consumption patterns. ‘Transpirational’ water requirements of vegetables are relatively large but do not provide reduction options because of nutritional constraints. The only option to reduce water requirements for food is to reduce the total amount of crops needed for a consumption pattern, for example, by the reduction of the use of animal feed, and thus smaller consumption of food items derived from livestock systems. The thesis assesses crop growth related water flows, or critical water requirements in agriculture, and improves the insight into the relationship between consumption and water requirements. The results have two important consequences for food policy issues. First, the thesis shows only small differences among water requirements for crops. Second, it shows that a doubling of food production does not necessarily imply a doubling of water use but only of ‘transpirational’ water use, a minor part of water requirements in agriculture. In some areas this amount is not available so that in order to meet increasing demand, yields in other areas will have to increase more than a factor of two. 77 ►► FFoooodd ccoonnssuummppttiioonn aanndd eeccoonnoommiicc ddeevveellooppmmeenntt,, aa ssppaattiiaall aanndd tteemmppoorraall ccoommppaarriissoonn In the coming decade, world population growth requires the production of more food. Along with population growth, many countries have shown increased purchasing power, causing not only per capita demand for more, but also for other food. Studies on human nutrition have shown that, throughout the world, a transition is taking place in which people shift towards the affluent food consumption patterns of western countries, characterized by large consumption of meat and oil, and relatively small consumption of staples. Such a transition causes increasing requirements for scarce natural resources, which makes it desirable to gain a better insight into factors that cause the transition. This thesis analyzes the relationship between annual GDP per capita, on the one hand, and food supply, the composition of food consumption, and the contribution of animal foods to nutritional energy, on the other. The thesis makes a spatial comparison of food consumption patterns of countries in various stages of economic development for the year 2001, and two temporal comparisons, a three century time trend for western Europe, and a forty year trend for southern Europe. All results show similar patterns of change. Total supply (kilocalories per capita per day) differs by a factor of two between low and high GDP. For low incomes, increase of GDP per capita parallels fast changes towards the per capita food supply levels of western countries, characterized by a large gap between supply and actual consumption. A second characteristic of the increase of GDP per capita and changes of consumption is an exchange of the fraction of nutritional energy from carbohydrates to fats and to animal foods, while the protein contribution to energy remains stable. People with low GDP derive nutritional energy mainly from carbohydrates, the contribution of fats is small, of protein similar to people with high GDP, and the contribution of animal sources negligible. People with high GDP derive nutritional energy mainly from carbohydrates and fats, the contribution of protein is similar to that of low GDP, and the contribution of animal sources is substantial. Whenever and wherever GDP increases occur, food supply, the composition of consumption, and the contribution of animal foods move in the direction of today’s affluent, western world food consumption patterns. The fastest changes occur in the lowest income categories, below 5000 1990 International Geary-Khamis dollars. The transition is completed at an annual GDP per capita of about 12 500 dollars. The findings have important consequences for food security. The European transition occurred in a gradual way, enabling agriculture and trade to keep pace with demand changes. A continuation of present economic trends not only causes increasing pressure on the food system, but also occurs faster than projections indicate. Especially economic development in Asia causes additional pressure on the global food system.

Summary ___________________________________________________________________________________

127

88 ►► PPaatthhwwaayyss ttoowwaarrddss ssuussttaaiinnaabbllee ffoooodd ccoonnssuummppttiioonn ppaatttteerrnnss Sustainable consumption includes a high quality of life, efficient use of natural resources, effective satisfaction of human needs, and equitable social development. One of the aims of this thesis is to indicate pathways towards sustainable food consumption patterns, as well as priority topics for future research. Differences in consumption patterns among countries as well as developments in time provide insight to increase sustainability, for example, by increasing efficiency, prevention, and substitution. The thesis shows practical pathways by comparing the use of land, ‘transpirational’ water, and energy for the affluent food consumption pattern of the Netherlands in 1990, by analyzing Dutch time trends over the period 1950-1990, and by comparing food supply and resource use for the Netherlands and Nigeria. Natural resource use is a combined effect of a consumption and a production subsystem. An analysis of the effect of consumption on natural resources, therefore, can only be carried out for a clearly defined production subsystem, generating relative results rather than actual resource use. The thesis makes comparisons of consumption systems using data from the Dutch production system of 1990. The results show that foods typical for an affluent consumption pattern, beverages, fats, animal foods, fruits, and sugar have substantial impacts on the use of natural resources. Especially the category of animal foods (meat, and dairy and eggs) has a large impact. In the Netherlands in 1990, it required 66 percent of total ‘transpirational’ water, 54 percent of agricultural land, and 45 percent of energy requirements. When a correction is made for developments in the production system, time trends show that actual, total resource requirements for the Dutch food consumption pattern over the period 1950-1990 showed different developments for the three natural reources. Although relative land requirements increased by 40 percent, actual, per capita land requirements decreased by 40 percent due to larger output per unit of land in agriculture over the period considered. Moreover, the thesis shows that actual ‘transpirational’ water requirements increased by 45 percent, and actual energy requirements doubled. For all resources, the effects of consumption changes were mainly related to consumption changes in the food categories of beverages (wine, beer, coffee), meat (beef, pork, poultry), and fats. For water, also larger consumption of fruits caused an increase of requirements. Affluent food consumption patterns have much larger resource requirements than patterns from developing countries, such as that of Nigeria. The latter shows relatively small requirements for all food categories, except for the category cereals, sugar, potatoes, vegetables, and fruits that seems inversely related to requirements for the other four. At present, developing countries still have relatively small natural resource requirements for food but growth potentials are large. For these countries, a transition towards sustainable food consumption patterns increases natural resource requirements. If differences in specific requirements among food items and categories are considered, these changes can take place in the most sustainable direction. For developed countries, changes that have occurred in the past decades provide insight into ways to promote more sustainable food consumption patterns with smaller resource requirements. Especially the large gap between per capita supply and actual consumption is an option to prevent resource use. This requires further research. For land, increased efficiency of land use by increasing yield levels is an option for reduction, although this requires larger energy input. Within food categories, substitution of foods by foods with similar functions but lower requirements is possible. This concerns substitutions of wine by beer, coffee by tea, and soyoil by sunflower or rapeoil. In the category of meat, requirements decrease in the following order: beef, pork, poultry; in the category of cereals, potatoes, vegetables and fruits, requirements for staples decrease in the order of rice, pasta, potatoes. Substitution among food categories is possible for food items of the category of meat, and dairy and eggs by protein-rich vegetal foods, such as peas. This option shows large resource reductions. The thesis studies some of the many factors from the economic and social dimension of sustainability that affect the environmental dimension through consumption. Although many relationships within the food system remain unquantified, the thesis contributes in a multidisciplinary way to priority topics on prevention and adaptation to human dimensions of environmental change. Future research should address resilience of the food system and focus on linkages among economic, social, and environmental systems. This type of research requires shared concepts and semantics on sustainability, collaboration efforts, and

Summary ___________________________________________________________________________________

128

commitment of scientific disciplines. In this respect, a priority place on the policy agenda is a prerequisite.

129

SSaammeennvvaattttiinngg 11 ►► IInnttrroodduuccttiiee Mensen hebben voedsel nodig om gezond te blijven en om niet van honger dood te gaan. Er bestaat echter een wezenlijk verschil tussen het voedsel van de mens en dat van dieren. Mensen eten namelijk niet alleen om in leven te blijven. Niemand eet of drinkt ‘zomaar iets’ met een willekeurige persoon op een willekeurig tijdstip. Het is een activiteit die in alle culturen aan regels en gewoonten is gebonden. Voor de productie, het transport, de verwerking en het bewaren van dit voedsel en de grondstoffen daarvan, zijn natuurlijke hulpbronnen nodig, waarvan landbouwgrond, zoet water en fossiele brandstoffen de belangrijkste zijn. Bijna alle landbouwgrond en het meeste zoete water dat door mensen wordt gebruikt, is bestemd voor de productie van voedsel. In westerse landen maakt de hoeveelheid energie voor voedsel 1/5 deel uit van het totale energieverbruik. Omdat voedsel een grote druk legt op schaarse natuurlijke hulpbronnen, is het van belang de relatie tussen de consumptie van voedsel en het gebruik van deze hulpbronnen in kaart te brengen. De belangrijkste doelen van dit proefschrift zijn: (i) het ontwerpen en ontwikkelen van meetmethoden voor duurzaamheid in voedselproductiesystemen; (ii) het analyseren van het effect van voedselconsumptiepatronen op natuurlijke hulpbronnen; en (iii) het in kaart brengen van mogelijkheden om voedselconsumptiepatronen te verduurzamen, rekening houdend met voedingskundige en culturele randvoorwaarden. Dit type onderzoek, dat de relatie tussen consumptiepatronen en duurzaamheid bestudeert, sluit aan bij de lange onderzoekstraditie van de IVEM, het Centrum voor Energie en Milieukunde, van de Rijksuniversiteit Groningen. De in dit proefschrift ontwikkelde kennis beoogt bij te dragen aan het duurzamer maken van bestaande voedselconsumptiepatronen, in zowel de westerse wereld als ook in ontwikkelingslanden. 22 ►► VVooeeddsseellssyysstteemmeenn Voedselsystemen bestaan uit twee subsystemen, een consumptie- en een productiesubsysteem. De hoeveelheid voedsel, die in een land nodig is, wordt bepaald door de grootte van de bevolking en de hoeveelheden en soorten levensmiddelen die mensen gebruiken: de voedselconsumptiepatronen. Voedselconsumptiepatronen verschillen sterk per land en veranderen in de tijd. In West-Europa trad na de Tweede Wereldoorlog een verschuiving op van de vrij sobere voeding van de jaren ’50 naar een meer luxe voeding in de jaren ’80 en ’90. In Zuid-Europa zette deze trend iets later in. Maar ook tussen welvarende westerse landen onderling bestaan er grote verschillen in de hoeveelheden levensmiddelen die door mensen worden geconsumeerd. Zo hebben de Zweden en de Nederlanders het hoogste koffieverbruik per persoon van Europa, en zijn de Ieren de grootste theedrinkers. Op wereldschaal bestaat het grootste verschil tussen de patronen van de arme ontwikkelingslanden, waar mensen zich vooral voeden met zetmeelrijk voedsel zoals cassave, en de rijke patronen in de westerse wereld. In sommige ontwikkelingslanden echter, zet een toenemende welvaart veranderingsprocessen in gang, waardoor voedselconsumptiepatronen opschuiven naar de patronen in de westerse wereld. Dit is bijvoorbeeld het geval in China en India, landen met een zo grote bevolking, dat veranderingen in consumptie een grote weerslag kunnen hebben op het gebruik van natuurlijke hulpbronnen. Voor de voedselproductie in de landbouw zijn ruimte, in de vorm van landbouwgrond, en zoet water nodig. In de 20e eeuw heeft de landbouw in westerse landen zich sterk ontwikkeld. Vooral na de Tweede Wereldoorlog gingen de opbrengsten per hectare zeer snel omhoog, vooral door een toename van het gebruik van kunstmest, gewasbeschermingsmiddelen en irrigatietechnieken, naast de introductie van nieuwe gewasvariëteiten. Dit betekent, dat het ruimtebeslag per kilo product is afgenomen. Om deze hogere opbrengsten te bereiken is wel meer (fossiele) energie nodig. Door een toename van de opbrengsten in de landbouw gaat dus het energiebeslag voor voeding omhoog. Zonder zoet water is landbouw onmogelijk. Het meeste water, dat door mensen wordt benut, wordt gebruikt voor irrigatiedoeleinden. Zoet water wordt echter steeds schaarser. Dit komt niet alleen doordat huishoudens en de industrie steeds meer water gebruiken, maar ook door het veranderen van neerslagpatronen, waardoor sommige landbouwgebieden verdrogen. Ruimte en water zijn nodig in de eerste schakel van de voedselproductieketen, de landbouw, energie is echter nodig in de hele keten: de landbouw, de industrie, de handel en het transport. Voor het energiegebruik maakt het veel uit waar en hoe een levensmiddel wordt geproduceerd en getransporteerd. Het beslag dat een levensmiddel legt op natuurlijke hulpbronnen is afhankelijk van de kenmerken van een bepaald productiesysteem. Het beslag dat een voedselconsumptiepatroon legt op natuurlijke

Samenvatting __________________________________________________________________________________________

130

hulpbronnen, is dan ook het resultaat van een specifieke combinatie van een productie- en een consumptiesubsysteem. 33 ►► OOnnttwweerrpp eenn oonnttwwiikkkkeelliinngg vvaann eeeenn mmeeeettmmeetthhooddee vvoooorr mmiilliieeuukkuunnddiigg dduuuurrzzaaaamm oonnddeerrnneemmeenn iinn ddee vvooeeddiinnggsssseeccttoorr Maatschappelijk verantwoord ondernemen, of duurzaam ondernemen, is voor bedrijven inmiddels een dagelijkse praktijk. De samenleving eist dat bedrijven, behalve aan economische belangen, ook aandacht besteden aan collectieve langetermijnbelangen op het gebied van welzijn en milieu. Duurzaam ondernemen kent drie aspecten: de sociale, de economische en de milieukundige. Dit proefschrift legt het accent op de milieukundige aspecten. Deze vinden veelal plaats op verschillende schaalniveaus, van lokaal naar globaal. De invloed van de individuele ondernemer wordt kleiner naarmate de milieubelasting zich afspeelt op een hoger niveau, zoals de effecten van kooldioxide-uitstoot op het klimaat. De noodzaak tot gedeelde verantwoordelijkheid van meerdere bedrijven, die een bijdrage leveren aan de totstandkoming van een eindproduct, wordt dan groter. Duurzaam ondernemen is echter lastig te meten, omdat er geen internationaal geaccepteerde standaard is over wat, wanneer, en hoe er moet worden gemeten en gerapporteerd. Bovendien maken bedrijven meestal deel uit van complexe productieketens, waardoor het mogelijk is dat milieukundige winst, behaald in het ene bedrijf, ten koste gaat van de milieuprestaties van een ander bedrijf in dezelfde keten. Het ontwerpen en ontwikkelen van een meetmethode voor duurzaam milieugericht ondernemen vergt dan ook een benadering, die het hele productiesysteem betrekt in de analyse. Er blijken al veel methoden te zijn ontwikkeld om duurzaam milieugericht ondernemen in het voedselsysteem te meten. Veelal wordt echter gekeken naar milieubelasting op lokaal schaalniveau, waarbij een breed scala aan indicatoren wordt gebruikt. Dit genereert grote hoeveelheden informatie, die nauwelijks bijdragen aan een beter inzicht in de duurzaamheid van een productiesysteem. Bovendien worden interacties in het systeem niet zichtbaar. Om belangrijke zaken met betrekking tot het milieu, die op wereldschaal spelen, zichtbaar te maken, stelt dit proefschrift voor om slechts drie indicatoren te gebruiken voor het meten van duurzaam ondernemen. Dit zijn het gebruik van de natuurlijke hulpbronnen land, zoet water en energie (van zowel fossiele als ook hernieuwbare energiedragers). De meetmethode wijst het gebruik in iedere stap van het productieproces toe aan het uiteindelijke product. Op deze manier kan de methode ook afwentelingen in het productiesysteem in kaart brengen. Het gebruik van de voorgestelde methode voor het meten van milieugericht duurzaam ondernemen in voedselproductieketens belicht zowel de milieugerichte prestatie van een individueel bedrijf, als ook die van een heel productiesysteem. De uiteindelijke uitkomst is het beslag van een levensmiddel op natuurlijke hulpbronnen. Bedrijven kunnen de informatie gebruiken voor het maken van profielvergelijkingen binnen het eigen bedrijf, maar ook tussen bedrijven of bedrijfsvestigingen. Consumenten kunnen met deze informatie het milieubeslag van levensmiddelen vergelijken. Hoewel deze methode is ontwikkeld voor milieugericht duurzaam ondernemen in de voedingssector, is zij toepasbaar voor vrijwel elke bedrijfssector. 44 ►► MMeetthhooddee vvoooorr ddee bbeerreekkeenniinngg vvaann hheett llaannddbbeessllaagg vvoooorr vvooeeddsseell De productie van voedsel vergt veel land, zowel akker- en tuinbouwland als grasland voor de veeteelt. Op wereldschaal vermindert de hoeveelheid land, die per persoon beschikbaar is voor voedsel. Zo nam tussen 1961 en 1998 dit areaal af van 1.5 hectare per persoon naar nog maar 0.8 hectare. De hoeveelheid land, nodig om een mens te voeden, hangt af van de combinatie van productiefactoren in de landbouw en een bepaald consumptiepatroon. Er zijn al veel studies gedaan naar voedselzekerheid, studies die vooral de mogelijkheden van de landbouw in kaart brachten om de groeiende wereldbevolking te voeden. De resultaten hiervan geven aan, dat toekomstige generaties kunnen worden gevoed gebruik makend van het bestaande landareaal. In die studies zijn de consumptiepatronen echter sterk vereenvoudigd en worden ze meestal uitgedrukt in beschikbare hoeveelheden graan. In het algemeen veroorzaakt toenemende welvaart een verandering van consumptiepatronen, ook voor voedsel. Om een beter inzicht te krijgen in de effecten van deze veranderingen op de hoeveelheid land, is gedetailleerde kennis nodig van het landbeslag van levensmiddelen. Hiermee kan het effect van veranderingen in de consumptie op het landbeslag van een geheel consumptiepatroon worden berekend. Dit proefschrift ontwikkelt een methode om het landbeslag van levensmiddelen en voedselconsumptiepatronen vast te stellen. Toegepast op het Nederlandse productiesysteem van 1990, een systeem met een zeer hoog renderende landbouw, levert dit het landbeslag van meer dan

Samenvatting

__________________________________________________________________________________________

131

honderd gangbare levensmiddelen op. (Zie Appendix A voor een volledig overzicht). Het landbeslag van levensmiddelen laat grote verschillen zien. Vooral dierlijke levensmiddelen als vlees en roomboter, maar ook plantaardige vetten, koffie en thee hebben een groot landbeslag (m2 per kg). Levensmiddelen waarvoor grondstoffen uit de akker- en tuinbouw nodig zijn, zoals brood, bier, suiker, aardappelen en groente, hebben een relatief klein landbeslag. Behalve het specifieke landbeslag van een levensmiddel, is ook de totale consumptie daarvan van belang. Een levensmiddel met een groot landbeslag waarvan maar weinig wordt gebruikt, zoals thee, draagt nauwelijks bij aan het landbeslag van een consumptiepatroon, terwijl een levensmiddel met een klein beslag maar met een hoge consumptie, zoals bier, een aanmerkelijke invloed heeft. Wanneer de gegevens van levensmiddelen worden gecombineerd met gegevens over consumptie, wordt duidelijk wat het effect is van consumptiepatronen op landbeslag. Zo had een Nederlands huishouden in 1990 het meeste land nodig voor vlees (29 procent), oliën en vetten (24 procent), zuivel en eieren (17 procent), en voor koffie, thee, wijn en bier (11 procent). Voor aardappelen, groente en fruit was slechts 5 procent nodig, en voor de overige levensmiddelen, waaronder brood, slechts 9 procent van het totaal. Het uiteindelijke landbeslag voor voedsel lag in Nederland in 1990 op ongeveer 0.2 hectare per persoon. De resultaten zijn kenmerkend voor de Nederlandse productiesituatie met hoge opbrengsten in de landen tuinbouw en in de veehouderij, en dus met een relatief klein landbeslag voor levensmiddelen. De methode is echter ook geschikt om andere combinaties van productie- en consumptiesystemen te bestuderen, zoals consumptiepatronen in andere landen of toekomstige ontwikkelingen. 55 ►► VVooeeddsseellccoonnssuummppttiieeppaattrroonneenn eenn llaannddbbeessllaagg Voedselconsumptiepatronen hebben een grote invloed op het landbeslag. Een beter inzicht in de effecten van deze veranderingen op het landbeslag is gewenst, vooral gezien de toenemende druk op de schaarse landbouwgrond. Voor de bestudering van de effecten van verschillen in consumptiepatronen op het landbeslag gebruikt dit proefschrift de hierboven beschreven methode, waarbij gegevens over landbeslag van levensmiddelen worden gecombineerd met gegevens over consumptie. Om nu de effecten van verschillen tussen voedselconsumptiepatronen op het landbeslag goed te kunnen bestuderen, gebruikt dit proefschrift de gegevens van slechts één productiesysteem, het Nederlandse van 1990. De resultaten van deze vergelijking geven dan ook niet het werkelijke landbeslag in vierkante meters weer, maar de relatieve verschillen tussen patronen. Voedselconsumptiepatronen verschillen sterk in tijd en plaats. Zo veranderde het gemiddelde patroon in Nederland sterk in de periode 1950-1990. Werd het patroon in 1950 gekenmerkt door een relatief lage consumptie van dure levensmiddelen als vlees, fruit, alcoholische dranken, koffie en kaas, en een relatief hoog verbruik van goedkopere levensmiddelen als aardappelen. In 1990 was de consumptie van dure levensmiddelen echter sterk toegenomen, terwijl dat van de goedkopere afnam. Hierdoor nam het relatieve landbeslag van het consumptiepatroon met 40 procent toe. Tussen de landen van de Europese Unie onderling waren in 1995 grote verschillen in consumptiepatronen te zien, resulterend in aanzienlijke verschillen in het landbeslag voor voedsel dat per persoon nodig is. Portugal had het kleinste landbeslag, Denemarken het grootste, ongeveer 40 procent meer. Wanneer gekeken wordt naar het landbeslag per consumptiecategorie, dan blijken de verschillen nog groter te zijn. De som van het kleinste landbeslag van de vijf verschillende consumptiecategorieën (dranken; vetten; vlees; zuivel en eieren; en granen, suiker, aardappelen, groente en fruit) verschilt een factor twee met de som van het grootste landbeslag. Het verschil tussen een karig menu, gebaseerd op tarwe, en een rijk menu zoals het Deense, is nog groter en verschilt een factor zes. Vooral in ontwikkelingslanden zijn voedselconsumptiepatronen nog erg eenvoudig en zijn ze vooral gebaseerd op zetmeelrijk voedsel. Wanneer deze landen zich gaan ontwikkelen, en overgaan op de rijke patronen die nu in westerse landen gebruikelijk zijn, zal het landbeslag sterk toenemen. Het effect van veranderingen in voedselconsumptiepatronen op de benodigde landarealen kan zelfs groter zijn dan dat van de groei van de wereldbevolking. 66 ►► VVooeeddsseellccoonnssuummppttiieeppaattrroonneenn eenn ddee bbeehhooeeffttee aaaann wwaatteerr Voor de productie van voedsel is veel zoet water nodig, ongeveer 70 procent van het totale menselijke verbruik ervan. Het is dan ook mogelijk dat, in sommige streken, in de nabije toekomst de vereiste toename van de landbouwproductie gepaard zal gaan met tekorten aan water. Er zijn al veel studies gedaan naar het waterbeslag van levensmiddelen. Het hanteren van verschillende systeemgrenzen in deze studies leidt echter tot onvergelijkbare resultaten. Om een beter inzicht te krijgen in de effecten van veranderingen in voedselconsumptiepatronen op de behoefte aan water is, net als voor land, gedetailleerde kennis nodig van het waterbeslag van levensmiddelen om zo het

Samenvatting __________________________________________________________________________________________

132

effect van veranderingen in de consumptie op het waterbeslag van een consumptiepatroon te kunnen berekenen. Landbouwkundige studies hebben modellen ontwikkeld om het waterverbruik in de landbouw te schatten. Dit levert praktische informatie op voor de dagelijkse praktijk. Dit proefschrift begint vanaf de andere kant. Op het moment dat er aan het einde van een groeiseizoen een oogst aanwezig is, is er voldoende water voor de groei van die oogst beschikbaar geweest. Gebruik makend van landbouwkundige kennis ontwikkelt dit proefschrift een methode om de groeigerelateerde waterbehoefte (transpiratiewater, nodig tijdens het gehele groeiseizoen) van gewassen te berekenen. Dit resulteert in de behoefte aan transpiratiewater voor gewassen (liters per kg versgewicht, liters per kg droge stof en liters per kJ geleverde energie). Omdat het watergehalte van gewassen sterk kan verschillen, verschilt ook de behoefte aan transpiratiewater sterk. Per eenheid droge stof zijn de verschillen kleiner, een factor twee. Dit verschil wordt vooral veroorzaakt door verschillen in chemische samenstelling tussen gewassen en door verschillen in de zogenaamde “harvest index”, de verhouding tussen de hoeveelheid economische oogst en de totale biomassaproductie. Wanneer de behoefte aan transpiratiewater wordt uitgedrukt in liters per eenheid energie, zijn de verschillen zo klein dat er geen mogelijkheden voor vervanging van gewassen bestaan om de behoefte aan water van een consumptiepatroon te verlagen. De behoefte aan water van groenten is relatief hoog, maar biedt geen mogelijkheid voor besparing, omdat ook rekening moet worden gehouden met voedingskundige randvoorwaarden. De enige mogelijkheid is dan de hoeveelheid gewassen, nodig voor een consumptiepatroon, te verminderen, bijvoorbeeld door een verbetering van de efficiëntie of door een lager verbruik van veevoeder met als resultaat een lagere productie en consumptie van levensmiddelen van dierlijke herkomst, zoals vlees en zuivel. Het proefschrift bepaalt de groeigerelateerde waterbehoefte (transpiratiewater) van gewassen. Dit blijkt slechts een klein deel te zijn van de totale waterbehoefte in de landbouw. Andere verliezen, zoals het verdwijnen van water naar diepere lagen, maken vaak een veel groter deel uit van de waterbehoefte. Dit betekent ook dat een verdubbeling van de productie in de landbouw per eenheid grond, mogelijk nodig voor het veiligstellen van de voedselzekerheid, niet inhoudt dat de totale waterbehoefte zal verdubbelen, maar slechts de behoefte aan transpiratiewater. In sommige streken zal deze hoeveelheid echter niet aanwezig zijn, zodat in andere streken de opbrengsten met meer dan een factor twee zullen moeten stijgen. 77 ►► VVooeeddsseellccoonnssuummppttiieeppaattrroonneenn eenn eeccoonnoommiisscchhee oonnttwwiikkkkeelliinngg In het komende decennium zal de voortgaande groei van de wereldbevolking de absolute vraag naar voedsel doen toenemen. Daarnaast zal in sommige landen een verhoging van het welvaartspeil een verschuiving veroorzaken in het type voedsel dat consumenten graag willen. Voedingskundige studies hebben aangetoond dat wereldwijd een transitie plaatsvindt naar de voedselconsumptiepatronen van westerse landen, gekenmerkt door een hoge consumptie van vlees en olie, en een lage consumptie van zetmeelrijk voedsel. Aangezien een dergelijke transitie het beslag op natuurlijke hulpbronnen zal doen toenemen, is het van belang deze te kwantificeren. Dit proefschrift legt het verband tussen het inkomen per hoofd (uitgedrukt in Bruto Nationaal Product) enerzijds, en de beschikbaarheid van voedsel (kilocalorieën per persoon per dag), de samenstelling van het voedsel (in macronutriënten), en de bijdrage van levensmiddelen van dierlijke oorsprong (in termen van het aandeel energie) anderzijds. Het proefschrift maakt een vergelijking voor een groot aantal landen in verschillende stadia van economische ontwikkeling verspreid over de hele wereld. Daarnaast maakt het proefschrift een tijdreeks van de ontwikkeling in West-Europa gedurende de laatste driehonderd jaar en een tijdreeks van de ontwikkeling in Zuid-Europa gedurende de laatste veertig jaar. De ruimtelijke en de historische analyses laten alle dezelfde kenmerken van de voedselconsumptiepatronen zien. De patronen van inkomensgroepen die minder verdienen dan 5000 International Geary-Khamis $ (G-K $) per jaar, kenmerken zich door kleine verliezen tussen beschikbaar voedsel en werkelijke consumptie, een relatief hoog aandeel van koolhydraten en een laag aandeel van vetten in de energievoorziening. Het aandeel van voedsel van dierlijke oorsprong in de totale energievoorziening is ongeveer drie procent. De voedselconsumptiepatronen van inkomensgroepen die meer verdienen dan 12 500 G-K $ per jaar kenmerken zich door grote verliezen tussen beschikbaar voedsel en werkelijke consumptie. De helft van het beschikbare voedsel wordt uiteindelijk niet opgegeten. De samenstelling van het voedsel wordt gekenmerkt door een relatief groot aandeel van vetten in de energievoorziening, ongeveer evenveel als van koolhydraten. De relatieve bijdrage van eiwit is ongeveer evenveel als die voor de consumptiepatronen van de lage

Samenvatting

__________________________________________________________________________________________

133

inkomens. De bijdrage van voedsel van dierlijke oorsprong aan de totale energievoorziening ligt tussen de 30 en 40 procent. De analyses laten zien dat veranderingen in inkomens per hoofd (in GDP per jaar) vooral voor de laagste inkomens (beneden 5000 G-K $) gepaard gaan met veranderingen in voedselconsumptiepatronen. Tussen de 5000 en 12 500 G-K $ neemt het tempo van verandering geleidelijk af, terwijl boven de 12 500 G-K $ geen verband meer kan worden aangetoond tussen veranderingen in het inkomen en de hoeveelheid beschikbaar voedsel per hoofd, de samenstelling van het patroon in macronutriënten en de bijdrage van voedsel van dierlijke oorsprong. De ontwikkelingen in Europa naar een voedselconsumptiepatroon met een hogere beschikbaarheid van voedsel, meer vetten, minder koolhydraten en meer voedsel van dierlijke oorsprong, was een geleidelijk proces dat de landbouw de mogelijkheid gaf zich aan te passen aan de veranderde vraag. Als de resultaten die in dit proefschrift gevonden zijn ook van toepassing zijn op die landen die het economische ontwikkelingstraject nog moeten doorlopen, betekent dit dat de druk op het landbouwsysteem om niet alleen meer, maar vooral ook anders te produceren sterk wordt vergroot. Vooral de economische ontwikkelingen in Azië gaan in de komende tien jaar een grote druk uitoefenen op het wereldvoedselsysteem. 88 ►► OOnnttwwiikkkkeelliinngg vvaann dduuuurrzzaammee vvooeeddsseellccoonnssuummppttiieeppaattrroonneenn Duurzame voedselconsumptiepatronen garanderen niet alleen een goede levenskwaliteit en de bevrediging van menselijke behoeften, maar ook zaken als een sociaal verantwoorde verdeling van voedsel en een efficiënt gebruik van natuurlijke hulpbronnen. Dit betekent dat de drie verschillende aspecten van duurzaamheid, de economische, de sociale en de milieukundige, in de definitie worden betrokken. Een van de doelstellingen van dit proefschrift is om bestaande voedselconsumptiepatronen duurzamer te maken. Hoewel de nadruk vooral ligt op het milieukundige aspect, zijn ook de twee andere van belang. De vergelijking van voedselconsumptiepatronen en de hiermee samenhangende druk op natuurlijke hulpbronnen laat grote verschillen zien tussen landen en grote verschillen in de tijd. Deze verschillen geven inzicht in manieren om duurzame patronen te ontwerpen. De vergelijking van het land-, transpiratiewater- en energiegebruik van een consumptiepatroon uit een welvarend ontwikkeld land, alsmede een tijdreeks van dit patroon, geeft inzicht in de mogelijkheden om voedselconsumptie duurzaam te maken. Dit geldt eveneens voor de vergelijking van een patroon uit een welvarend land en een ontwikkelingsland. Voor die analyse vergelijkt het proefschrift de consumptie van Nederland met die van Nigeria, een ontwikkelingsland. Aangezien het gebruik van natuurlijke hulpbronnen afhangt van het type productiesysteem, gebruikt het proefschrift voor deze vergelijking gegevens van het Nederlandse systeem in 1990. Een duurzaam voedselconsumptiepatroon is mogelijk door verhoging van de efficiëntie bij productie en consumptie, preventie en substitutie. Het blijkt dat levensmiddelen kenmerkend voor een welvarend patroon: dranken, vetten, levensmiddelen van dierlijke herkomst, fruit en suiker, een groot beslag leggen op natuurlijke hulpbronnnen. Vooral de invloed van vlees, zuivel en eieren is groot. In 1990 was in Nederland 66 procent van het transpiratiewater, 54 procent van het land en 45 procent van het energiegebruik van voedsel nodig voor de categorie levensmiddelen van dierlijke herkomst. Wanneer een tijdreeks van het werkelijke gebruik van natuurlijke hulpbronnen wordt gemaakt, waarbij ook rekening wordt gehouden met veranderingen in het productiesysteem, dan blijkt dat in Nederland tussen 1950 en 1990 het werkelijke gebruik van land per hoofd met 40 procent afnam, het gebruik van transpiratiewater met 40 procent toenam, en het energiegebruik zelfs verdubbelde. Voor alle natuurlijke hulpbronnen gold, dat vooral de toenemende consumptie van dranken (wijn, bier en koffie), vlees (vooral varkens- en kippenvlees) en vetten, de druk op de hulpbronnen deed toenemen. Voor water gold bovendien een toename van de druk door een grotere fruitconsumptie. De voedselconsumptiepatronen van ontwikkelde landen veroorzaken per persoon een veel grotere druk op natuurlijke hulpbronnen dan patronen uit ontwikkelingslanden zoals Nigeria. Het gemiddelde patroon in dat land laat voor vier van de vijf consumptiecategorieën (dranken; vetten; vlees; zuivel en eieren) een relatief klein gebruik van natuurlijke hulpbronnen zien, terwijl alleen het gebruik voor de categorie granen, suiker, aardappelen, groente en fruit relatief groot is. Ontwikkelingslanden hebben vergeleken met ontwikkelde welvarende landen een relatief klein gebruik van natuurlijke hulpbronnen. Het groeipotentieel is echter groot. Wanneer ook rekening wordt gehouden met de sociale dimensie van duurzaamheid neemt in ontwikkelingslanden de consumptie van levensmiddelen, kenmerkend voor een welvarende levensstijl, toe en gaat dus ook de druk op natuurlijke hulpbronnen omhoog. Wanneer rekening wordt gehouden met het specifieke beslag dat individuele levensmiddelen leggen op het gebruik van natuurlijke hulpbronnen, dan kan voor deze ontwikkeling de meest duurzame

Samenvatting __________________________________________________________________________________________

134

manier worden aangewezen. Voor de ontwikkelde landen bieden de veranderingen in de afgelopen decennia een leidraad om het voedselconsumptiepatroon te verduurzamen. Vooral het grote verschil tussen de beschikbaarheid per hoofd van voedsel op nationaal niveau en de hoeveelheid voedsel die daadwerkelijk door een individu wordt gegeten, is een indicatie dat de efficiëntie, waarmee met voedsel om wordt gegaan aanmerkelijk kan worden vergroot. Dit vergt echter verder onderzoek. Voor de hulpbron land is een grotere efficiëntie in de landbouw een manier om het landgebruik terug te dringen. Dit vergt evenwel meer energie. Binnen consumptiecategorieën kan het gebruik van hulpbronnen worden teruggedrongen door substitutie door levensmiddelen met een gelijkwaardige functie, maar met een kleiner beslag op deze hulpbronnen. Dit betreft bijvoorbeeld de substitutie van wijn door bier, koffie door thee, en sojaolie door zonnebloem- of raapolie. Voor vlees neemt het beslag op hulpbronnen af in de volgorde rundvlees, varkensvlees, kippenvlees. Voor zetmeelrijke levensmiddelen neemt het beslag af in de volgorde rijst, deegwaren, aardappelen. Ook is het theoretisch mogelijk dierlijke levensmiddelen te vervangen door eiwitrijke plantaardige levensmiddelen, zoals peulvruchten. Ook hierdoor neemt het beslag op hulpbronnen af. Een goed inzicht in het beslag dat individuele levensmiddelen en levensmiddelencategorieën leggen op het gebruik van natuurlijke hulpbronnen, is een hulpmiddel bij het richting geven aan de ontwikkeling van duurzame voedselconsumptiepatronen. Het proefschrift richt zich op enkele van de vele factoren die een rol spelen in het voedselsysteem. Deze factoren hebben gemeen dat ze deel uitmaken van economische en sociale systemen en via consumptie het milieusysteem beïnvloeden. Hoewel veel verbanden binnen het voedselsysteem niet in kaart worden gebracht, levert het multidisciplinaire karakter van dit proefschrift een beter inzicht in milieukundige problemen op wereldschaal. Het geeft verder mogelijkheden aan om de druk op natuurlijke hulpbronnen te verminderen. Toekomstig onderzoek zou zich moeten richten op de kwetsbaarheid en het aanpassingvermogen van voedselsystemen. Met name de verbanden tussen de economische, de sociale en de milieukundige aspecten van duurzaamheid zijn daarbij van belang. Dit type onderzoek vergt vergaande samenwerking tussen verschillende wetenschappelijke disciplines, en eenduidige definities en opvattingen met betrekking tot duurzaamheid. Een belangrijke plaats op gezamenlijke politieke agenda’s is daarbij vereist.

135

CCuurrrriiccuulluumm vviittaaee

Winnie Gerbens-Leenes was born in Groningen, the Netherlands in 1953. In 1971, she graduated from secondary school in Heerenveen where she received the certificate athenaeum-B. Winnie continued her study at the School for dietetics “De Schutse” in Nijmegen. After an interval in which she raised her two children and worked for a marketing agency, she started to study at the Open University (OU) in 1991. In 1999, she finished her MSc study in environmental sciences at the OU with a thesis on the effects of Dutch household food consumption on land and energy resources. Her research was performed at the Center for Energy and Environmental Studies IVEM of the University of Groningen. During the period 1999 to 2000, she was employed as a researcher at the IVEM in which she wrote the “Groene Kookboek” (Green Cooking Book), a translation of scientific knowledge into practical information for energy and land reduction potentials for household food consumption. From 2000 to 2006, she was employed at the IVEM to conduct a Ph.D. study on natural resources for food. In 2004, she carried out a research at the International Institute for Applied Systems Analysis (IIASA) in Vienna, Austria on transitions in food consumption patterns in Africa towards western patterns and consequences for land and energy requirements. Winnie has published most of her research in international, peer reviewed scientific journals.

136

137

LLiisstt ooff ppuubblliiccaattiioonnss

Gerbens-Leenes, P.W., 1999. Indirect ruimte- en energiebeslag van de Nederlandse voedselconsumptie. IVEM-onderzoeksrapport nr. 102. University of Groningen, the Netherlands. Gerbens-Leenes, P.W., 1999. Consumptiepatronen beïnvloeden ruimte- en energiebeslag voor voeding. Arena, mei 1999, pp. 18. Gerbens-Leenes, P.W., 2000. Groen Kookboek. IVEM-onderzoeksrapport 103. University of Groningen, the Netherlands. Nonhebel, S., Gerbens-Leenes, P.W., 2001. Resource use efficiencies in primary production systems: comparison between two disciplines. In: Ulgiati, S., Brown, M.T., Giampietro, M., Herendeen, R.A., Mayumi, K., (editors). Proceeding of the 2nd International Workshop Advances in Energy studies, exploring supplies, constraints and strategies. Porto Venere, Italy, May 23-27th, 2000. Pp. 235-242. Steg, L., Vlek, C., Feenstra, D., Gerbens-Leenes, P.W., Karsten, L. Kok, R. Lindenberg, S., Maignan, I., Moll, H., Nonhebel, S., Schoot Uiterkamp, T., Sijtsma, T. and van Witteloostuijn, A., 2001. Towards a comprehensive model of sustainable corporate performance. Three-dimensional modelling and practical measurement. University of Groningen-The Netherlands. Department of Economics, Environmental Sciences, Management Science, Psychology and Sociology. Interim report University of Groningen, the Netherlands. Gerbens-Leenes, P.W., Nonhebel, S., Ivens, W.P.M.F., 2002. A method to determine land requirements relating to food consumption patterns. Agriculture Ecosystems & Environment 90, 47-58. Gerbens-Leenes, P.W., Nonhebel, S., 2002. Consumption patterns and their effect on land required for food. Ecological Economics 42, 185-199. Vlek, C.A.J., Steg, E.M., Feenstra, D., Gerbens-Leenes, P.W., Lindenberg, S., Moll, H.C., Schoot Uiterkamp, A.J., Sijtsma, F., Witteloostuijn van, A., 2002. Een praktisch model voor duurzaam bedrijfspresteren. Economisch Statistische Berichten 87, 524-527. Gerbens-Leenes, P.W., Nonhebel, S., 2002. Food consumption patterns and the land required for the production of this food. In: Changes at the other end of the chain; Everyday consumption in a multidisciplinary perspective. Butijn, C.A.A., Groot-Marcus, J.P., Linden van der, M., Steenbekkers, L.P.A., Terpstra, P.M.J., (editors). Shaker Publishing BV, Maastricht, the Netherlands. Gerbens-Leenes, P.W., Nonhebel, S., Moll, H.C., 2002. Energy use in food production chains: the consequences of reduction strategies. In: Paradoxes in food chains and networks. Trienekens, J.H., Omta, S.W.F. (editors), Wageningen Academic Publishers, the Netherlands. P.W. Gerbens-Leenes, P.W., 2002. Indicators for corporate environmental responsibility in the food production chain. IVEM working paper May 2002-1. Center for Energy and Environmental Studies (IVEM), University of Groningen, the Netherlands. Gerbens-Leenes, P.W., Moll, H.C., Schoot Uiterkamp, A.J.M., 2003. Design and development of a measuring method for environmental sustainability in food production systems. Ecological Economics, 46 pp 231-248. Gerbens-Leenes, P.W., Nonhebel, S., 2003. Voedselconsumptiepatronen beïnvloeden behoefte aan landbouwgrond. Voeding Nu 9, 28-29. Gerbens-Leenes, P.W., Nonhebel, S., 2003. Het Groene Kookboek voor energiezuinige maaltijden. Voeding Nu 12, 22-23. Gerbens-Leenes, P.W., Schilstra, A.J., 2004. An historic perspective on the vulnerability of the Netherlands to environmental change: life in a cut-away. In: Wise use of Peatlands, 12th International peat congress 6-11 June 2004, Tampere, Finland.

List of publications __________________________________________________________________________________________

138

Gerbens-Leenes, P.W., Nonhebel, S., 2004. Food and land use. The influence of consumption patterns on the use of agricultural resources. In: Eating and drinking. Trends and tensions after 1945. Proceeding elfde sociaal wetenschappelijke studiedagen SISWO, Amsterdam, 22-23 april 2004. Gerbens-Leenes, P.W., Nonhebel, S., 2004. Critical water requirements for food, methodology and policy consequences for food security. Food Policy 29, 547-564. Gerbens-Leenes, P.W., Nonhebel, S., 2005. Food and land use. The influence of consumption patterns on the use of agricultural resources. Appetite 45, 21-31. Gerbens-Leenes, P.W., Moll, H.C., 2005. Measurement of Environmental Sustainability of Food Companies using a life Cycle Approach. In: Conference proceedings 11th Annual International Sustainable Research Conference, Helsinki, Finland, 6-8 June 2005. Gerbens-Leenes, P.W., Nonhebel., S., 2005. Dietary Trends and Land use, the impact of changing food consumption patterns on agricultural land resources. In: Proceedings 33td international Conference Industrial Ecology for a Sustainable Future. June 12-15 2005 Stockholm, Sweden. Pp. 136-138. Gerbens-Leenes, P.W., 2006. Resource use for food. The case of changing food consumption patterns and the use of natural resources in Nigeria. IVEM-onderzoeksrapport 114. University of Groningen, the Netherlands. Gerbens-Leenes, P.W., Nonhebel., S., 2006. Food consumption and economic development, a spatial and temporal comparison. Submitted to Food Policy.

139

AAcckknnoowwlleeddggeemmeennttss

Chapter 2 of this thesis was part of a multidisciplinary project on the scientific modeling and measuring of Sustainable Corporate Performance (SCP) aimed at the development of a three-dimensional model of SCP involving economic, social and environmental dimensions. The project was supported by the University of Groningen Ubbo Emmius Fund and partly financed by the Royal Ahold Corporation in Zaandam, the Netherlands. I thank the Ubbo Emmius Fund and Ahold for their financial support. The stimulating project coordination by Charles Vlek and Linda Steg of the University of Groningen is greatly acknowledged. Most of the work presented in this thesis has been published in scientific journals. I want to thank my coauthors, Wilfried Ivens, Henk Moll, Sanderine Nonhebel and Ton Schoot Uiterkamp for their contributions. Beside comments of several anonymous reviewers on earlier drafts of my papers, I wish to thank, Mario Giampietro for his comments on the paper that formed the basis for Chapter 3, and John Lintott for his comments on the paper for Chapter 4. I also want to thank my supervisor Mahendra Shah of the International Institute for Applied Systems Analysis (IIASA). Part of the work I performed under his supervision formed the basis for Chapter 7 of this thesis. The work of this thesis has also been an input for the FIDES (Food Information as a Determinant for Eliciting Sustainability) project, funded by the ‘Verantwoorde Voeding Programma’ of ZonMW, part of NWO. Finally, I wish to thank the members of the external reading committee, Prof. dr. J. Van Andel, Prof. dr. K. Blok and Prof. dr. R. Leemans for their effort.

Documents

University of Groningen Natural resource use for food Leenes ...Wiskunde en Natuurwetenschappen aan de Rijksuniversiteit Groningen op gezag van de Rector Magnificus, dr. F. Zwarts,