future european agricultural landscapes—what can we learn from existing quantitative land use...
TRANSCRIPT
Future European agricultural landscapes—What can we learn from
existing quantitative land use scenario studies?
Gerald Busch *
Center for Environmental Systems Research, University of Kassel, Kurt-Wolters-Str. 3, Germany
Available online 18 January 2006
Abstract
The structure of agricultural production and spatial patterns of agricultural land use in Europe are expected to face major changes over the
next decades due to changes in global trade, technology, demography and policies. This paper presents a set of 25 scenarios comprising
information on quantitative land use changes in Europe. The scenarios have been selected from studies with different foci, operating on both
different spatial scales and different time horizons. Given the diversity of quantitative scenarios this review illustrates the scenario design and
its quantification, and evaluates the results of land use/cover changes on a European level. Major gaps of current scenario exercises and
suggestions for improvement are topics of the discussion section (Section 4).
The focus of this review is on the comparison of selected driving forces and on agricultural land use/cover change in ‘‘Western Europe’’
(i.e. EU-15 plus Switzerland and Norway for some studies). Results show large differences in future land use/cover changes ranging from
moderate decreases (15%) to large increases (30%) depending on the assumptions about global trade, increase in agricultural productivity and
biofuel production. Domestic demand is a minor factor of land use/cover change since population is only changing slightly, and the
consumption level is stable and decoupled from economic growth. Scenarios show that the rate and direction of land cover change differ over
time. Considerable shifts towards grassland abandonment in many scenarios reflect the changes in agricultural management. Increasing
biofuel production as a result of both increasing energy demand and pro-active climate policies takes up considerable areas in many scenarios
and prevents substantial abandonment of agricultural land.
Although comparable quantitative results concerning European agricultural land use/cover change are only available on a very aggregated
level, the results are important to be dealt with when discussing future challenges of rural areas.
# 2005 Elsevier B.V. All rights reserved.
Keywords: European land use scenarios; Agricultural land use/cover; Land cover change
www.elsevier.com/locate/agee
Agriculture, Ecosystems and Environment 114 (2006) 121–140
1. Introduction
Agriculture has shaped many European landscapes over
centuries. This has given rise to unique semi-natural
environments with a rich variety of habitats and species
dependent on the continuation of farming. Agricultural land
use still dominates the European landscape since it covers
about 45% by area of the EU-25 states. However, the extent
of agricultural land is declining and the value added to the
* Present address: Bureau for Applied Landscape Ecology and Scenario
Analysis, Am Weißen Steine 4, 37085 Gottingen, Germany.
Tel.: +49 1212 315 916666; fax: +49 1212 624 124124.
E-mail address: [email protected].
0167-8809/$ – see front matter # 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.agee.2005.11.007
overall annual gross domestic product (GDP) of the EU-25
countries is merely 2% (EUROSTAT, 2005).
In the last 50 years, a considerable change of agricultural
production has taken place. Technological progress and the
aim to establish nationally and internationally competitive
agricultural production have produced a marked intensifica-
tion and specialization in agriculture supported by the
Common Agricultural Policy (CAP), a strong European
policy framework with various incentives and subsidies
indirectly affecting not only agricultural land use but the
rural landscape (Commission of the European Communities,
2000; European Environment Agency, 2003; European
Communities, 2004). Agriculture plays a key role in the
management of natural resources in rural areas and
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140122
agricultural policy is increasingly perceived by regional
stakeholders and politicians as an integrative part of rural
development complementing other sectoral policies (Bal-
dock et al., 2002; European Communities, 2004) and taking
into account the multiple functions of rural areas (e.g.
recreation, ecosystem services, agricultural production,
economic infrastructure). Thus, increasing migration away
from rural areas, a decrease of agricultural employees, aging
of farmers and abandonment of both farm holdings and
agricultural land have been a reason for major concerns.
With the implementation of both the Agenda 2000 in
1999 (European Commission, 1999) and the Lisbon strategy
in 2001 (European Commission Internet document, 2005),
two important steps were taken to integrate agricultural
policy with rural development issues in the European Union.
The relaunched Lisbon strategy in 2005 aims at combining
and modifying existing programs, such as the LEADER + -
initiative (Liaisons Entre Actions de Developpement de
l’Economie Rurale) or the INTERREG-III and SAPARD
programs (Special Accession Programme for Agriculture &
Rural Development) to foster rural development (i.e.
diversification of production, innovation, increasing R&D
measures) which in turn is a key element of restructuring the
agricultural sector (Baldock et al., 2002; European Com-
munities, 2003, 2004).
Over the next decades many regions in Europe will have
to face major demographic changes, structural changes in
agricultural production and the diversification of the
regional economy (European Commission, 2004). How
will European agricultural land use evolve in time and what
are the major uncertainties in future agricultural production?
Changes in agricultural land use and land cover reflect
economic causes, policy measures as well as spatial
planning objectives and show a wide range of impacts,
including biophysical and socio-economic changes and
feedbacks between land use and its drivers. Scenario
generation is an appropriate tool to develop plausible visions
of future pathways of European land use. Scenarios can help
to understand how critical uncertainties will play out and
what new factors will come into play and can, thus, result in
surprising and innovative insights (Davis, 2002). Literature
review on European land use scenarios revealed that a large
number of different scenario exercises exist (Busch et al.,
2004), but only a limited number of studies address driving
forces and land use/cover changes quantitatively—which is
the entry point of this review paper. Thus, comprehensive
and instructive studies, such as the ‘‘VISIONS’’ exercise
(Rotmans et al., 2001) or the ‘‘Scenarios Europe 2010’’
(Bertrand et al., 2001) could not be considered because they
focus on qualitative information about European land use.
Though quantitative and comprehensive, the ‘‘Ground for
Choices’’ study (Wetenschappelijke Raad voor het Reger-
ingsbeleid, 1992) as well as the ‘‘European Mid-Term
Review’’ exercise (Commission of the European Commu-
nities, 2002) were not considered either. The ‘‘Ground for
Choices’’ study is outdated because the scenarios that had
been developed for this study refer to the European
agricultural situation in 1990 and more than half of the
scenario period from 1992 until 2015 has elapsed. The
CAPRI modeling (Britz and Heckelei, 1997) for the
‘‘European Mid-Term Review’’ shows a very short time
horizon (2002–2009) and was assessed to be a projection of
current policies rather than a scenario building exercise.
Seven environmental studies comprising quantitative
scenarios on agricultural land use/cover change in Europe
have been selected by searching publicly available
information. This review focuses on agricultural land use
and land cover changes, and evaluates the different
assumptions (i.e. the direct and indirect driving forces
behind the scenarios) made in these scenarios. Five of the
seven studies reviewed are global in their scope and only two
studies comprise both the national and regional level in their
analysis of driving forces.
In consequence, the aim of this paper is two-fold: (1) to
illustrate the level of information which can be derived from
currently available quantitative land use scenarios and (2) to
evaluate if the scenario design and its quantification
adequately elaborate on problems of European agriculture
being currently discussed (European Environment Agency,
2003; European Commission, 2004, 2005; European
Communities, 2004). The goal of this review is to discuss
scenario results of seven studies with respect to quantitative
changes in agricultural land us/cover change on a European
level. The comparison of agricultural land use and land
cover change in this study covers ‘‘Western Europe’’ (i.e.
EU-15 countries plus Norway and Switzerland for some
studies).
2. Material and methods
2.1. Overview of the review approach
To evaluate the quantification of scenario conditions a
scheme as presented in Fig. 1 is followed. The main
qualitative scenario characteristics of the seven studies
selected are introduced (a), and then the quantification of
population growth and Gross Domestic Product as the two
most important exogenous drivers of land use/cover change
are briefly described (b). The major commonalities and
differences of the models being used for quantification of
land use/cover change simulation are discussed (c). Based
on a selected set of driving forces the outcome of land use/
cover change modeling concerning both changes in arable
land and pasture area (d) is discussed in the results section.
2.2. Scenario study selection
Twenty-four studies have been reviewed based on
publicly available information (i.e. scientific papers,
published reports, the Internet and technical documents)
in order to select quantitative land use scenarios being
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140 123
Fig. 1. Schematic representation of the review approach. Sources: Raskin et al. (1998), IMAGE-Team (2001), Organisation for Economic Co-Operation and
Development (2001), Kemp-Benedict et al. (2002), Raskin et al. (2002), Gallopın and Raskin (2002), United Nations Environment Programme (2002), Potting
and Bakkes (2004), EURuralis (2004), Schroter (2004), Klijn et al. (2005), van Meijl et al. (2006) and Rounsevell et al. (2005, 2006).
spatially explicit, covering at least the EU-15 states and
having a time horizon of more than 15 years ahead.
Table 1 shows the seven studies that were selected. These
seven studies comprise 25 scenarios with a time horizon
ranging from 2020 to 2100. Due to different finalization of
the scenario studies, the reference year (base year) varies
between 1995 and 2002. Further, the scenario studies cover
different spatial scales. The first five studies shown in
Table 1 are global in their scope and address driving forces
on a global scale and for different world regions (e.g.
Northern America and Western Europe). The two latter
studies focus on Europe taking into account global level
information but comprising both the national and regional
level in their analysis of driving forces.
With a focus on environment, poverty reduction and
human values, the scenarios developed by the Global
Scenarios Group (GSG Futures) are based on various
exercises carried out at the Stockholm Environment
Institute-Boston (Raskin et al., 1998, 2002; Gallopın and
Table 1
Studies included in the review
Base year Time horizon No. of scenarios Global Spatia
World
Global Scenarios Group Futures/Stockholm Environment Institute Boston (GSG
1995 2050 4 X X
Special Report on Emission Scenarios/National Institute for Public Health and t
1995 2100 4 X X
Global Environmental Outlook 3/National Institute for Public Health and the En
2002 2032 4 X X
Global Environment Outlook 3/Stockholm Environment Institute Bostona (GEO-
2002 2032 4 X X
Environmental Outlook/OECD (OECD)
1995 2020 1 X X
Advanced Terrestrial Ecosystem Analysis and Modeling/University of Louvain-l
2000 2080 4 X X
EURuralis/Wageningen University/National Institute for Public Health and the E
2000 2030 4 X X
a The two studies use the same storylines, but different tools for quantificatio
Raskin, 2002). The GEO-3 scenarios are part of the third
Global Environmental Outlook coordinated by UNEP and
described in the GEO-3 report (United Nations Environment
Programme, 2002). The four scenarios have an environmental
focus considering the social and economic spheres. Building
to a large extent on the GSG-Futures scenarios, regional
policies are more elaborated in the SEI-Futures scenarios. In
both studies, the four scenarios address fundamentally
different societal visions which are characterized by (a)
essential continuity with current patterns, (b) fundamental but
undesirable societal change and (c) fundamental and
favorable societal transformation, respectively.
The IPCC Special Report on Emission Scenarios
(SRES) focuses on greenhouse gas emissions assuming
that policies to mitigate emissions are not implemented
(Intergovernmental Panel on Climate Change, 2000). The
IMAGE-SRES scenarios referred to in this study represent
an elaboration of the IPCC-SRES scenarios (IMAGE-
Team, 2001). The four scenarios build on the gradient of
l scale Focus
regions National Sub-national
Futures)
Environment, society
he Environment (SRES)
Climate
vironment (GEO-3/RIVM)a
Environment
3/SEI)
Environment
Environment, economy
a-Neuve (ATEAM)
X X Environment
nvironment (EURuralis)
X X Rural environment
n and were thus considered separately.
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140124
two major dimensions: whether the emphasis is on
economic wealth (A) or on sustainability and equity (B)
and whether the world is globally oriented (1) or regionally
focused (2).
The OECD Environmental Outlook (Organisation for
Economic Co-Operation and Development, 2001) reflects an
economy-based vision of environmental impacts. The
scenario of the OECD Environmental Outlook relies on
the ‘‘reference’’ variant of the ‘‘Bending the Curve’’
scenarios, developed by the Global Scenarios Group (Raskin
et al., 1998; Gallopın and Raskin, 2002), which in turn
addresses the vision of essential continuity with current
patterns.
The focus of the ATEAM study lied on the assessment of
European Ecosystem vulnerability. Within an integrative
assessment approach four scenarios for socio-economic
development and land use change were developed. Based on
the four global SRES scenarios, a European specific
interpretation of major driving forces was elaborated taking
into account European and national policy measures.
The main goal of the EURuralis exercise was to develop
different visions about the future of the EU-25 rural areas
taking into account possible effects on the environment, the
economy and the socio-cultural sphere. Drawing from the
SRES and GEO-3 scenario approaches, the EURuralis
project developed four scenarios which build on the gradient
of two major dimensions: (1) ranging from a world which is
facing increasing globalization to regionally oriented
economies and cultural blocks and (2) ranging from low
government regulation to ambitious governance with high
regulation (EURuralis, 2004).
2.3. Qualitative scenario information and scenario
classification
All scenarios reviewed refer to key elements acting as
driving forces, such as demography, culture and society,
economic development, technology, policy and governance
and environment. The qualitative information derived from
this study review is visualized in Fig. 2, in order to address
the major differences and commonalities of the scenarios.
The scenarios considerably differ in the aspects of solidarity,
environment and policy regulation. Further, the scenarios
diverge into globally oriented and regionally focused
pathways. Population growth, technological innovation
and economic growth show different but positive dynamics.
Many of the driving forces show similar patterns of change,
which is not surprising since all 25 scenarios build on two
scenario exercises: the SRES scenarios and the GSG
scenarios, which were then elaborated and interpreted.
Based on the information shown in Fig. 2, a gradient of
two major dimensions is constructed: (1) whether the
emphasis is on self-interest or on solidarity and (2) whether
the world is globally oriented or regionally focused. As a
result of this matrix approach, four scenario categories are
designated (1) Global Markets, (2) Global Society, (3)
Continental Barriers, and (4) Regional Sustainability. The
results shown for selected driving forces and land use/cover
change in Section 3 refer to these four scenario categories
(see Fig. 3).
2.4. Quantification of population and economic growth
as major driving forces
Demographic and economic assumptions play a funda-
mental role in driving demand for agricultural products and
thus in changes of agricultural land use/cover. An overview
on data used to quantify population growth and gross
domestic product is given in Table 2. Both indicators were
used as exogenous drivers in all scenario studies. Concern-
ing population data, four studies referred directly or
indirectly to projections developed by the United Nations.
These projections comprised, however, different variants
and each study made use of its own downscaling or regional
aggregation procedure (Center for International Earth
Science Information Network, 2002). The other three
studies, which used population data from the Phoenix
model (Hilderink, 2000) took the macroeconomic data from
the WorldScan model, an economic multi-region, multi-
sector, applied general equilibrium (AGE) model (Nether-
lands Bureau for Economic Policy Analysis, 1999). The SEI
studies made use of Worldbank projections (Raskin et al.,
1996) and developed own interpretations based on these
projections (Raskin et al., 1998). The OECD utilized data
from its own organization for the Environmental Outlook
study, which were based on the work by Burniaux (2000).
2.5. Modeling agricultural land use/cover change
Given the qualitative information on important driving
forces, socio-economic boundary conditions and policy
decisions from the scenarios, land use models were used to
project how much land is utilized where and for what
purpose. Since the scenario studies had different foci, and
operated on both different spatial scales and different time
horizons, distinct techniques were applied to compute land
use/cover change. Quantification of land use/cover change,
however, followed a general scheme: land use/cover change
is a result of changing demands of agricultural products,
production technology and biophysical suitability. Food
demand, economic growth, international trade and policies
drive the demand of agricultural products. Demand for
agricultural production leads to land use requirements. The
land use requirements result in land use/cover changes
depending on production technology, biophysical suitability
and spatial restrictions of land resources.
2.5.1. Modeling approaches used for quantification
Both the GSG Futures scenarios study and the GEO-3
study carried out by the Stockholm Environment Institute
used the PoleStar model for quantification (Kemp-Benedict
et al., 2002). PoleStar is a so-called accounting model
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140 125
Fig. 2. Major global driving forces and their qualitative patterns of change.
(Potting and Bakkes, 2004) combining exogenous eco-
nomic, resource and environmental information on a global
and regional level (Fig. 4).
To calculate food supply and demand and the resulting
land use/cover changes, PoleStar started with human
dietary and industrial demands for agricultural products.
These demands for agricultural production were translated
into requirements for land, water and nutrient inputs.
International trade of food products was based on current
patterns of food trade. Changes in production technology
were addressed by interpretation of current trends or
borrowing data from other studies (Food and Agriculture
Organization, 2003). Land allocation was computed on a
continental level. Land use competition and land
conversion was addressed by using statistical information
on potential agricultural land and conversion rules,
respectively.
The SRES scenarios we refer to in this review were
computed with IMAGE. In contrast to PoleStar, IMAGE is a
dynamic integrated assessment modeling framework for
global change (Alcamo et al., 1998; IMAGE-Team, 2001).
IMAGE endogenously computed the demand for land,
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140126
Fig. 3. Classification of scenarios.
agricultural productivity and land use competition, and
provided spatially explicit data for 17 world regions
simulating land use/cover change on a half-by-half degree
grid. An iterative optimization between food demand and
land use requirements was calculated taking into account
income, biophysical suitability of land and technology.
Trade was introduced by exogenously prescribing self-
sufficiency ratios for each of the 17 world regions. Land
use/cover change was computed pixel-based taking into
account neighborhood to other cells and applying spatial
allocation rules including a hierarchy of land use types
(Strengers, 2001).
As shown in Fig. 4, the IMAGE model was used for an
alternative quantification of the GEO-3 scenarios (Potting
and Bakkes, 2004). Thus, it allowed comparing quantifica-
Table 2
Sources used for projections of population and economic development
GDP
GSG Raskin et al. (1996, 2002)
SRES Worldscan (Netherlands Bureau for Economic Policy
Analysis, 1999)
GEO-3/RIVM Worldscan (Netherlands Bureau for Economic Policy
Analysis, 1999)
GEO-3/SEI Raskin et al. (1996, 2002)
OECD Burniaux (2000)
ATEAM Center for International Earth Science
Information Network (2002)
EURuralis Worldscan (Netherlands Bureau for Economic
Policy Analysis, 2003)
tion results from different models within one scenario
framework.
For the OECD Environmental Outlook a soft-link between
two models was established. The JOBS model (Organisation
for Economic Co-Operation and Development, 2001), a
neoclassical general equilibrium model, was used to calculate
the sectoral demands, prices and commodity production (e.g.
crops, livestock and forest products) on a global level and for
different world regions. In JOBS, computation of agricultural
commodity demand and production was iteratively processed
in order to reach equilibrium between supply and demand.
International trade was described as substitutability between
domestic products and foreign products. Exogenous assump-
tions concerning overall agricultural productivity growth,
agricultural management factors and land supply were used
Population
United Nations (2000), Raskin et al. (2002)
Phoenix (Hilderink, 2000)
Phoenix (Hilderink, 2000)
United Nations (2000), Raskin et al. (2002)
United Nations (2000)
Center for International Earth Science
Information Network (2002)
Phoenix (Hilderink, 2000; Netherlands Bureau for Economic
Policy Analysis, 2003)
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140 127
Fig. 4. Modeling tools used and spatial scales addressed for quantification of land/use cover change.
(Organisation for Economic Co-Operation and Development,
2001). The PoleStar System then used the economic and
demographic variables as input to calculate the quantitative
land use/cover changes.
The ATEAM study utilized the European agricultural
demand figures from the IMAGE 2.2 model as basic input
for agricultural demand–supply modeling to take care of
both domestic demand of agricultural goods and global trade
patterns including agricultural demand from other world
regions. Depending on the European-specific interpretation
of the SRES scenarios, the IMAGE numbers on agricultural
demand were modified and accompanied by own calcula-
tions of major driving forces (e.g. population growth and
economic growth). The Louvain-la-Neuve land cover model
(LLN-LCM) computed land use/cover change spatially
explicit on a 10 min � 10 min grid scale. Productivity
changes were taken into account exogenously (Ewert et al.,
2006), and spatial allocation rules were applied to compute
land use patterns (Rounsevell et al., 2005, 2006) and
included European policy options.
EURuralis used the most complex approach to simulate
land cover changes on different spatial scales by linking an
economic model to a biophysical model (van Meijl et al.,
2006). Changes in agricultural land were computed by using
combined results of the general economic equilibrium
model GTAP (Hertel, 1997) and the integrated assessment
model IMAGE. GTAP computed changes in agricultural
production, and demand for agricultural land at national to
global level taking into account international trade. GTAP
calculated productivity increase, management factors and
land supply endogenously. Further the implementation of an
endogenous quota mechanism improved the reflection of
European agricultural policies. The IMAGE model used
information on yields, feed efficiency rates and demand for
land from GTAP to calculate agricultural productivity and
demand for land. Through re-iteration the land projections
converged and produced a harmonization of both biophy-
sical and economic land use processes (Klijn et al., 2005;
van Meijl et al., 2006). The information on land use/cover
change at national level was used by the CLUE-S model to
downscale land use demands to land use patterns at a
1 km � 1 km grid scale resolution (Verburg et al., 2006).
2.5.2. Geographical coverage of the models
For the comparison of land use scenarios on a European
level, it is important to define a common geographical
coverage. Due to different study foci and separate models
used for quantification the geographical coverage of world
regions is, unfortunately, not consistently addressed. Europe
does not exist as one single world region but is addressed by
sub-regions, which are, again, defined differently. Each study,
however, includes the EU-15 countries in its regional
classification. The differences due to classification addressing
four additional countries in some studies are assessed not to be
notably important for this review because the agricultural area
in these countries is very small. This review refers to the
regional classification given as ‘‘Western Europe’’.
2.5.3. Drivers of land use/cover change
Major driving forces of land use/cover change as
addressed by the models are shown in Fig. 5. The first
part of Fig. 5 depicts the drivers of demand for agricultural
production. Demand for agricultural production is a
combination of domestic demand and demand from other
world regions. All models addressed this issue by computing
variables, such as population growth, diets, caloric intake,
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140128
Fig. 5. Drivers of agricultural land use/cover change.
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140 129
economic development, policies and international trade.
Major differences occurred in the consideration of demo-
graphic and economic variables. Changes in age structure
and migration were only addressed in the ATEAM and the
EURuralis study. Economic effects of tariffs and export
subsidies were only considered in the OECD and the
EURuralis studies. Further, the GSG and the GEO-3/SEI
studies do not address biofuel crop demand.
The second section of Fig. 5 addresses agricultural
production and its drivers. All models computed both crop
and animal production but with different detail. All models
considered major factors of technology and management but
the number of crops and animals simulated, varied
considerably. In the IMAGE-GTAP framework productivity
of 11 crop types was simulated endogenously. PoleStar only
addressed three major crop categories (wheat, rice, other
crops). The LLN land cover model used exogenously
calculated productivity of wheat as a proxy for food crop
productivity. In addition to food crops, grass and fodder
crops, and three biofuel crop types were considered.
Calculation of animal production is most advanced in the
IMAGE-GTAP modeling framework in terms of animal
types being considered.
In the IMAGE-GTAP modeling framework the imple-
mentation of policy measures on a European or sub-
European level was most elaborated. Economic indicators
were only computed by JOBS and GTAP. Spatial restric-
tions, biophysical constraints and land use transition
pathways were not addressed by PoleStar. Land use/cover
classification varied considerably between only five classes
in the LLN land cover model and 21 classes in the IMAGE
calculations. Comparison of land use/cover changes is
hampered because different classifications are used.
2.6. Driving force selection for quantitative comparison
To discuss land use/cover change results of the different
studies, eight driving forces have been selected. The criteria
for the driving force selection have been data availability and
Table 3
Driving forces used for the quantitative comparison
Driving force
Demand
Income
Population
Caloric intake
Production technology
Fertilizer input
Crop yield
Grass and fodder production
Trade and production
Crop demand
Crop production
Animal demand
Animal production
representation of important factors of land use/cover change,
which are: (1) food demand, (2) production technology and
(3) production and trade.
Table 3 shows the eight driving forces selected for
comparison. The ratio of agricultural demand to agricultural
production shows how much of a commodity is consumed
regionally, and thus indicates the export-orientation of
agricultural production.
3. Quantitative comparison of driving forces and
agricultural land cover change
The quantitative comparison is based on the scenario
classification given in Section 2.3 and the selection of
driving forces described in Section 2.6. Since the base years
of the studies vary, the geographical coverage is not entirely
consistent, and the time horizon being considered ranges
from 2020 to 2050, the data are normalized to the base year.
The figures presented in the following sections show relative
percent changes. The changes of the driving forces are
calculated as average annual percentage changes and are
presented as a comparison of two driving forces in each
figure. The diagonal in the driving force figures represent an
equivalent growth rate (1:1 line) of both indicators.
Three categories of agricultural land use/cover changes are
compared: (1) cropland, (2) pasture and (3) abandoned land.
All figures depict the scenario studies in the same order,
which is: (1) GSG, (2) SRES, (3) GEO-3/SEI, (4) GEO-3/
RIVM, (5) OECD, (6) ATEAM and (7) EURuralis. Note,
that not all studies are represented in each scenario category.
3.1. Drivers of land use requirements
3.1.1. Demand
With the selected set of driving forces (population
growth, income, crop demand, animal demand, and caloric
intake), changes in European demand for agricultural
production are illustrated. Starting with the two important
Explanation
Growth in gross domestic product per capita
Total population growth
Intake of food calories per capita
Synthetic fertilizer input per hectare of cropland
Crop yield per hectare
Green fodder and legume production
Regional demand for food and feed crops
Total crop production (food and feed)
Regional demand for animal products
Total animal production
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140130
Fig. 6. Average annual percentage change in agricultural demand of selected indicators.
exogenous variables population growth and economic
growth, all scenarios assume a continued economic growth
in Europe and only small population changes in ‘‘Western
Europe’’ (Fig. 6). Economic growth shows a higher
variability and ranges from slight increases in the ‘‘Regional
Sustainability’’ category to moderate increases in the
‘‘Global Markets’’ category. Caloric intake, however, is
decoupled from economic growth in all scenarios since the
current level of food consumption and caloric intake in
‘‘Western Europe’’ is high.
Most scenarios show both an increasing crop demand and
an increasing animal demand. Increasing crop demand being
considerably higher than population growth reflects increas-
ing resource consumption in form of food demand. Small
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140 131
changes in caloric intake, however, indicate that the increase
in animal and crop demand is only partially caused by
raising domestic demand and a shift in diets. For most
scenarios, the increase in demand for animal products is of
the same magnitude as the increase in population, showing
that diets are fairly stable.
Within the ‘‘Global Society’’ category, the GSG ‘‘Great
Transition’’ and the GEO-3/SEI ‘‘Sustainability First’’
scenarios reflect a special pathway showing a considerable
decline in crop demand and demand for animal products
(Fig. 6d and f). Decreasing per capita food demand (Fig. 6k),
and a shift in diets towards less meat consumption (Fig. 6m)
reflect changing consumer preferences in the ‘‘Regional
Sustainability’’ category. In contrast, ‘‘Security First’’,
shows both a considerable increase in caloric intake
(Fig. 6h) and a shift in diets towards more meat consumption
(Fig. 6i).
3.1.2. Production technology
Eight variables are used to highlight changes in
agricultural production technology that is influencing land
use/cover change (Fig. 7). Increasing yields, increasing
production and intensification of agricultural production
dominate the overall picture of agricultural production
technology (Figs. 7 and 8). In many scenarios, decreasing
green fodder production accompanied by increasing animal
production indicate both further intensification of livestock
production as well as a shift in species towards poultry and
pork which heavily rely on grain feed.
High rates of innovation are an attribute of the ‘‘Global
Markets’’ scenarios category and are expressed as high
increases in crop yields which mainly stem from better
management and biotechnological development rather than
from increased fertilizer input (Fig. 7a). The assumptions of
increase in crop productivity differ considerably and the
widespread combination of crop yields and fertilizer
application indicate significant differences in agricultural
management. In both the ‘‘Global Society’’ category and the
‘‘Regional Sustainability’’ category, a strong focus on
technological improvements and better management is
reflected by increasing crop productivity while fertilizer
input decreases in all scenarios. In contrast to the scenarios
of the ‘‘Global Markets’’, the shift in agricultural manage-
ment results from strong environmental regulation. Animal
production (Figs. 7e and 8e), however, shows a similar
pattern of intensification and shift in production as described
for the ‘‘Global Markets’’ scenarios. Cropland area
decreases due to intensified crop production and increasing
crop yields combined with stable or decreasing demand for
cop products (Fig. 7m). Pasture area declines since livestock
production is intensified and a shift towards grain-based
fodder consumption is assumed. Note that total animal
production is decreasing in the ‘‘Regional Communities’’
scenario of the EURuralis study, but pasture areas are
maintained due to extensive grazing management (Fig. 8k)
supported by tariffs on agricultural imports (van Meijl et al.,
2006). In the ‘‘Continental Markets’’ category, the two GSG
scenarios show a distinct management of agricultural
production. A considerable, fertilizer-based intensification
of crop production is accompanied by an expansion of
grazing management and a lower shift towards pork and
poultry production, resulting in an extension of green fodder
production (Fig. 7g and h).
3.1.3. Production and trade
Both the relation of crop and animal production to crop
and animal demand are used to illustrate the trade
orientation of agricultural production (Fig. 8). In the three
scenario categories ‘‘Global Markets’’, ‘‘Global Society’’
and ‘‘Continental Barriers’’ crop and animal production by
far exceeds the domestic demand in most scenarios
reflecting global food demand stemming from other world
regions. This development is most pronounced in the
scenario categories representing global and open markets.
In three scenarios of the ‘‘Global Society’’ category,
animal production only serves the domestic demand
(Fig. 8f). In the EURuralis scenario, trade liberalization
and abolished domestic support payments (van Meijl et al.,
2006) cause considerably reduced crop production (Fig. 8e).
In contrast the export-orientation of crop production in the
SEI scenarios is even higher than in the scenarios of the
‘‘Global Markets’’ category since environmental policies in
developing regions prevent uncontrolled expansion of
agricultural land and thus trigger additional crop imports
from other regions (Kemp-Benedict et al., 2002; Gallopın
and Raskin, 2002). In the ‘‘Continental Barriers’’ scenario
category demand for food imports from other world regions
is assumed to increase since world population is consider-
ably increasing and regional barriers hamper technological
diffusion. Subsidies and domestic support are assumed to be
typical policies and thus support European farmers to
produce for the world market. Fig. 8h and i illustrate this
export-oriented agricultural production of many scenarios.
Three of six scenarios, however, follow a more regionally
oriented pathway in crop production showing a nearly
balanced situation of domestic crop demand and crop
production (Fig. 8h). In contrast to the other scenario
categories agricultural production in the ‘‘Regional Sustain-
ability’’ category only meets the domestic demand. Since
there is no export-oriented production, crop production
increases only slightly and animal products demand is even
assumed to decrease (Fig. 8l and m).
3.2. Land use/cover change
3.2.1. Cropland change
Assumptions on production technology, domestic
demand and international trade determine the extension
of agricultural land. In three of four scenario categories, the
various combinations of these three elements result in both
considerable decrease and substantial increase of cropland
area (Fig. 9). In the ‘‘Regional Sustainability’’ scenario
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140132
Fig. 7. Average annual percentage change in production technology of selected indicators.
category, decreasing cropland area is the only outcome
caused by less food exports combined with decreasing
domestic demand and increasing productivity.
In the ‘‘Global Markets’’ category, the SRES and the
ATEAM scenarios denote the extremes in cropland area
change (Fig. 9a). The SRES scenario shows, surprisingly,
only a marginal increase in crop productivity but an even
higher gain in total crop production, and thus, results in a
substantial expansion of cropland area to 2050. The increase
in cropland area does not only relate to export-oriented food
crop production but also results from extended biofuel
production (see Fig. 10) because of very high energy
demand in an economically oriented future (IMAGE-Team,
2001). The ATEAM scenario represents the opposite
combination, i.e. crop yield increase outweighs by far the
rise in crop production, which in turn causes a considerable
decline in cropland area to 2050. This pathway is enforced
by crop allocation on optimal location, which is an
interpretation of the global and economic scenario orienta-
tion (Rounsevell et al., 2005). A similar effect appears in the
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140 133
Fig. 8. Average annual percentage change in agricultural production and trade of selected indicators.
scenario of the EURuralis study. Increase in crop yields
and crop production nearly balances (Fig. 7a and c) but
cropland declines. This is caused by abandonment of less
productive land due to abolished subsidies and reduced
domestic support payments ((Klijn et al., 2005; van Meijl
et al., 2006).
In the ‘‘Global Society’’ scenario category, the diver-
gence of decreases and increases in cropland area is smaller
than shown for the scenarios of the ‘‘Global Markets’’
category. The IPCC-SRES ‘‘B1’’ scenario and the GEO-3/
RIVM ‘‘Sustainability First’’ scenario show an increase of
cropland area (Fig. 9b) due to highest crop production
combined with lowest increase in crop yields (Fig. 7d and f).
The two scenarios with moderate to high increases in yields
but reduced growth in crop production (ATEAM ‘‘B1’’ and
EURuralis ‘‘Global Co-operation’’) show a decrease in crop
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140134
Fig. 9. Change in cropland area.
area. Both the GSG ‘‘Great Transition’’ scenario and the
GEO-3/SEI ‘‘Sustainability First’’ scenario assume a
transition from a growth-oriented to a sustainability-oriented
world. In the ‘‘Continental Barriers’’ category, again,
different combinations of crop yield increase, domestic
demand and export-oriented production result in both a
considerable decrease of cropland and moderate to high
growth (Fig. 9c). In the ‘‘Regional Sustainability’’ category
increasing crop yields combined with decreasing crop
production result in a continuous decline of cropland area, as
shown for the ATEAM scenario and the EURuralis scenario
in Fig. 9d. The SRES scenario indicates a transition from
cropland decline towards a slight increase in cropland area
reflecting agricultural dis-intensification.
3.2.2. Pasture change
Diverging pathways of pastureland are characteristic for
all scenario categories. In the two global-oriented cate-
gories, the development of pastureland varies between slight
and considerable decreases in area. In the two regional-
oriented categories, the divergences ranges between slight
increases and considerable decreases in pasture area. The
rate at which pastureland decreases and its trend reflect both
the different assumptions on intensification in livestock
production (e.g. increasing productivity, less grazing) and a
shift from grass-based production systems towards fodder
crops. This shift in turn results from two changes in animal
production (1) change in feed composition and (2)
preference change in meat consumption from ‘‘red meat’’
(beef) towards ‘‘white meat’’ stemming from pork and
poultry production (Sere and Steinfeld, 1996).
In the ‘‘Global Markets’’ category, the change in feed
composition is only marginal since the shift from green
fodder to feed crops is assumed to be less than 3% in all
scenarios (IMAGE-Team, 2001; Kemp-Benedict et al.,
2002; Bouwman et al., 2005). Hence, a combination of a
considerable increase in livestock productivity, a shift in
consumer preferences towards pork and poultry meat
consumption, and changes in agricultural management
towards stable-based livestock production cause a decrease
in pastureland. In the ‘‘Global Society’’ category, minor
decreases in the ATEAM ‘‘B1’’ scenario and the EURuralis
‘‘Global Co-operation’’ scenario reflect assumptions on
grassland preservation as part of both environmental policy
goals and restrictive planning. In the ‘‘Continental Barriers’’
category the extensification of grassland production with a
shift towards pasture-based fodder production cause a slight
increase of pasture area in the EURuralis scenario and the
GSG scenarios. In the ‘‘Regional Sustainability’’ scenario
category, only the EURuralis scenario shows slightly
increasing pasture area resulting from preservation of
grazing management.
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140 135
Fig. 10. Change in pasture area.
3.2.3. Extension of agricultural land and biofuel
production
Fig. 11 summarizes the development of extension in
agricultural land and biofuel production. Expansion of
agricultural land is mainly caused by international food
demand and increasing biofuel crop production. The
expansion of agricultural land is most pronounced in the
global-oriented scenarios with large food exports. Abandon-
ment is a result of increasing agricultural productivity and a
shift in agricultural production systems. The decrease of
agricultural land is most pronounced in scenarios with lower
food exports and less biofuel production.
In the ‘‘Global Markets’’ category, increase in biofuel
production is not a result of pro-active environmental
policies but a consequence of a fast increasing high energy
demand and an energy mix given by external models
containing renewables (IMAGE-Team, 2001). In the
‘‘Global Society’’ category, increase in biofuel area is
reduced in the ATEAM and SRES scenario due to lower total
energy demand. In contrast, the implementation of climate
mitigation policies enhances biofuel production in the GEO-
3/RIVM scenario. Increasing biofuel production in the
‘‘Continental Barriers’’ category does not reflect a higher
environmental awareness but is part of the security-oriented
self-reliance policy expressed in the general scenario
characteristics (see Section 2.3). Note that in the SRES
and in the GEO-3/RIVM scenario biofuel production takes
place on former pasture areas or on other land converted to
agricultural land. The ATEAM scenario is the only scenario
showing abandonment of agricultural land stemming from
both cropland areas and pastureland. Without the substantial
production of biofuels, agricultural land would diminish by
about 40% in area. In the ‘‘Regional Sustainability’’ scenario
category, biofuels are promoted as a regional and more
sustainable energy carrier since a major focus in this
scenario category is on environmental and socio-economic
sustainability. In the ‘‘Regional Communities’’ scenario,
biofuel crops are produced on abandoned cropland whereas
in the ATEAM ‘‘B2’’ and the SRES ‘‘B2’’ scenarios biofuel
crops grow on former pasture areas. All three scenarios show
a pronounced increase in area for biofuel production
compared to the base year situation.
4. Discussion
4.1. Storylines and scenario building
A large set of quantitative scenarios is available that
describe changes in European agricultural land. All 25
scenarios reviewed, however, referred to two global scenario
sets explored by the IPCC and the SEI. This is not a
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140136
Fig. 11. Change in agricultural land and biofuel production.
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140 137
drawback since these scenario sets reflected the major
uncertainties (globalization versus regionalization, solidar-
ity versus self-interest, economic-orientation versus envir-
onment orientation), and described a broad range of future
pathways. All scenario studies reviewed followed a top-
down approach taking into account actors and factors on a
global and European level. The ATEAM and the EURuralis
further reflected actors and factors on a national and sub-
national level and integrated sectoral perspectives in order to
address land use/cover changes in rural areas. Even in
entirely liberalized markets and a globalized economy,
farmers who act on a local level take proximate decisions.
The top-down approaches lack a feedback between
decisions taking place on a local/regional level and the
European level. To keep track of important decision levels
top-down and bottom-up approaches have to complement
one another. Both narrative storylines and quantitative
modeling have to address this issue. Thus, narrative
storylines have to refer to regionally specific spatial
processes. Such a scenario building process should make
use of two elements: (1) iteration between qualitative and
quantitative information and (2) participatory approaches.
Computation of regional spatial patterns could be used to
check the consistency of qualitative scenario information.
Re-iteration could help to harmonize qualitative and
quantitative scenario information. Participatory scenario
approaches would help to reflect the regional complexity of
European land use processes in the scenario building process
(European Commission, 2004). The PRELUDE exercise
carried out by the European Environment Agency was a first
attempt to develop participatory land use scenarios for
Europe and to combine the qualitative information and the
model quantification in an iterative way (Mc Glade, 2004).
Insights from this study could be a starting point for further
integration between existing scenario information and more
region-specific participatory approaches.
4.2. Quantification of land use/cover change
Given the qualitative information on major uncertainties
of the scenario storylines, the quantification of the important
driving forces revealed that technological development,
agricultural management and demand from other world
regions emerge as key factors of changes in European
agricultural land use/cover change. Changes in domestic
food demand were assumed to be of minor importance since
both population and caloric intake were not significantly
increasing. Agricultural land use continues to be a dynamic
process with multiple plausible pathways in all scenarios.
Non-linear land use/cover changes indicate that it is
important to examine different time scales to understand
the future consequences of changing trends. Major changes
in land use/cover are caused by intensification of agricultural
land use, abandonment and conversion of both natural
vegetation and pasture to cropland. These diverging
pathways imply major challenges for regional and European
policies since massive changes in socio-economic infra-
structure in rural areas could be expected and substantial
environmental impacts would have to be addressed. The
information provided by current scenario exercises could help
to stimulate the policy discussion of future rural development.
Substantial increases in agricultural productivity and a focus
on optimal location as assumed in some scenarios pose the
question what kind of rural landscapes we want to maintain in
Europe. Could land simply be abandoned and become natural
vegetation after decades or do we want to maintain patchy
landscapes with high recreation and aesthetic value? What are
possible strategies to cope with structural changes in
agricultural production assuming a highly industrialized
agriculture and shift in consumer preferences as indicated by
some scenarios? Do we want to create multifunctional
landscape all over Europe or could we benefit from functional
landscapes in both environmental and economic terms? How
will be dealt with the trade-off between agricultural expansion
due to agricultural demand from other world regions and
environmental protection? Will the economic return of
agricultural production justify substantial changes of land-
scape composition? Is biofuel production an appropriate
alternative to abandonment or can it even justify conversion of
natural vegetation?
Current scenario exercises provide stimulating informa-
tion on European land use/cover change but there are some
limitations and drawbacks that need to be highlighted. Five of
the seven studies reviewed are global in their perspective.
They address land use change only on a European level. These
global studies have a broad environmental perspective, but
land use changes are commonly viewed from a sectoral
perspective addressing changes in agriculture, forestry and
urban areas separately. A comparison of the different studies
is hampered due to land cover classification inconsistencies,
various spatial definitions of Europe, and limited spatial
resolution of quantitative results. In consequence, it was only
possible to compare relative changes in land use/cover
distribution on a European level lacking the interpretation of
changing land use/cover patterns.
Within the same scenario category the assumptions on
major driving forces and land use/cover change varied
considerably. Population growth and income as important
exogenous variables differed notably since data from various
sources were applied. On the one hand, this is positive since it
highlights the uncertainties of the scenario exercises – a very
important information scenarios can provide – on the other
hand a comparison of scenario results is hampered because of
too many variables. Current scenario exercises lack a
transparent approach how qualitative scenario information
is translated into numbers. As an example, the interpretation
of the same qualitative information resulted in a notably
different quantification for the GEO-3 scenarios (GEO-3/
RIVM and GEO-3/SEI), which in turn led to significantly
different land use/cover changes. A transparent methodology
and documentation would be of great added value for the
interpretation of quantitative modeling results.
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140138
International trade negotiations and European-specific
agricultural policies were only marginally and indirectly
addressed in the globally focused studies. Market intervention
through European agricultural policies was adequately
addressed only in the EURuralis study because policy
measures, such as domestic support, and tariffs were explicitly
implemented. The ATEAM study reflected European policy
intervention in the agricultural sector only through broad
assumptions about productivity changes, changes in over-
supply and different spatial distribution of agricultural area.
Current agricultural practices in Europe differ strongly
both between regions and within regions reflecting regional
socio-economic and biophysical conditions (European
Environment Agency, 2003; European Communities, 2003;
European Commission, 2005). Land use changes reflect the
decision making process by those who control land resources
(Verburg et al., 2006). In consequence, simulation of land use
change has to address the spatial scale and the decision
making level of the main actors. Depending on the scenario
philosophy, this decision making level could vary from the
global to the local level or be a combination of different levels.
In all scenarios, the proximate actors of both agricultural land
use and rural development are local farmers and regional
institutions. Hence, the quantification of driving forces on a
national or even a European level does not adequately reflect
this issue. Regional information has to be implemented in
both the scenario building process and its quantification.
Apart from the spatial scale or level of information, the
studies lack important information on land use intensity and
rural development. Transition into organic farming systems
or changes in farming structure need to be addressed since
both have an important impact on agricultural landscapes.
Aging of farmers and migration from rural to urban areas are
important factors (European Environment Agency, 2003;
European Communities, 2003; European Commission,
2005), which need to be addressed in assessments of
changing rural landscapes. Agricultural land use (change) is
increasingly perceived as multi-level, multi-actor and multi-
domain process (Knickel and Kok, 2003; European
Commission, 2004). Thus, modeling of agricultural change
needs to consider different levels, actors and domains.
4.3. Modeling framework
Scenario quantification in the different studies was
carried out with a limited set of modeling tools for a very
simple reason: only very few models can generate
quantitative long-term projections of future land use/cover
changes at regional to global levels. In consequence,
different scenarios borrow from the same quantitative data
computed by the same models and diminish their qualitative
diversity. Hence, scenario development and interpretation
may be hampered by model limitations.
The discussion of the quantitative results turns out that
changes in European agricultural land use/cover changes are
only adequately addressed when taking into account
different levels of information and being spatially explicit
on a high-resolution grid. In the ATEAM and the EURuralis
study these goals were, at least partly, achieved by (soft-)
linking different models. A potential problem of both
approaches is repeated computing of input- and output
information from different models without cross-checking
the underlying processes. The ATEAM and the EURuralis
study applied land use information from the IMAGE model
for downscaling procedures to create spatial patterns of land
use/cover changes. The ATEAM study utilized the European
agricultural demand figures from the IMAGE 2.2 model as
input information for land use/cover change modeling
(Rounsevell et al., 2005, 2006). The information on
agricultural demand in turn was calculated by IMAGE
using, e.g. biophysical suitability and crop yields (IMAGE-
Team, 2001). In the ATEAM study crop yields and
management factors were adopted from other external
sources to compute land use/cover changes. In the
EURuralis project CLUE-S used information on spatial
extension of land use changes from IMAGE to create spatial
patterns of land use/cover change (Verburg et al., 2006).
Variables, such as soil texture, temperature, precipitation
and slope were used to compute biophysical suitability in the
downscaling process. IMAGE computed agricultural areas
with similar variables but from different sources and with a
distinct spatial resolution. Hence, in both studies agricultural
land was re-allocated without checking possible contra-
dictions between modeling approaches.
Another potential problem appearing in these kinds of
downscaling procedures is a gap between level of
information and spatial resolution. Some driving force
information is only available at national or European level
but the spatial output information on land use/cover change
in case of the EURuralis study is 1 km � 1 km. A validation
based on time slices of European land use data would help to
answer the question if the current level of input information
is appropriate to generate patterns of land use/cover change
on a high-resolution grid.
5. Conclusion
The currently available quantitative scenario information
on European land use/cover change indicates that both
European agriculture and rural areas might undergo a process
of radical change in the next decades. Providing various
pathways of future development the scenario studies could
stimulate policy discussion with respect to possible implica-
tions of future agricultural demand, agricultural productivity
and impacts on rural areas. The ATEAM and the EURuralis
studies for the first time linked continental level information
with downscaling procedures and thus allowed to analyze
regional patterns of land use/cover change.
The current approaches of top-down modeling should be
accompanied by region-specific bottom-up modeling to
strengthen the value of quantitative scenario studies in
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140 139
supporting policy discussion. The different spatial patterns
resulting from various trade-offs between land use types are of
particular interest for regional environmental assessments and
can only be addressed by quantitative modeling with a high
spatial resolution. European-specific land use scenarios with a
strong spatial and regionally varied perspective would be a
major step ahead in addressing this issue. To reach this goal of
comprehensive regional land use scenarios, more detail need
to be taken into account with respect to multiple functions of
landscapes, different intensities in land use, land use-relevant
policies, regional land use competition and the impacts of
urban and infrastructure development on rural landscapes.
Acknowledgements
I would like to thank Eric Kemp-Benedict, Bas Eickhout
and Isabelle Reginster for their data support. I would like to
thank Mark Rounsevell and Joseph Alcamo for interesting
discussions about land use scenarios. Finally, I would like to
thank the PRELUDE project team for providing inspiring
insight into participatory scenario approaches.
References
Alcamo, J., Leemans, R., Kreileman, E., 1998. Global Change Scenarios of
the 21st Century. Results from the Image 2.1 Model. Pergamon, Oxford.
Baldock, D., Dwyer, J., Vinas, J., 2002. Environmental integration and the
CAP. A Report to the European Commission, DG Agriculture. Institute
for European Environmental Policy, Brussels.
Bertrand, G., Michalski, A., Pench, A., 2001. Scenarios for Europe 2010.
Five possible futures for Europe. Working Paper. European Commission
Forward Studies Unit, Brussels.
Bouwman, A.F., van der Hoek, K.W., Eickhout, B., Soenario, I., 2005.
Exploring changes in world ruminant production systems. Agric. Syst.
84 (2), 121–153.
Britz, W., Heckelei, T., 1997. Pre-study for a medium-term simulation and
forecast model of the agricultural sector for the EU. Institute for
Agricultural Policy, University of Bonn, Bonn.
Burniaux, J.-M., 2000. A multi-gas assessment of the Kyoto Protocol.
Working Papers, 270. OECD Economics Department, Paris.
Busch, G., Lubkert, B., Alcamo, J., 2004. Future landscapes: a review of
European scenarios about land use and environment. Internal Rport for
the 1. Stakeholder Meeting of the PRELUDE Poject. European Envir-
onmental Agency, Copenhagen.
Commission of the European Communities (CEC), 2002. Mid-term review
of the common agricultural policy: communication from the Commis-
sion to the Council and the European Parliament, 10.7.02. COM, 2002
(394). Commission of the European Communities, Brussels.
Commission of the European Communities (CEC), 2000. Communication
from the Commission to the Council and the European Parliament:
Indicators for the integration of environmental concerns into the Com-
mon Agricultural Policy. Commission of the European Communities,
Brussels.
Center for International Earth Science Information Network (CIESIN),
2002. Country-level GDP and Downscaled Projections Based on the A1,
A2, B1, and B2 marker scenarios, 1990–2100. CIESIN, Columbia
University, NY.
Netherlands Bureau for Economic Policy Analysis (CPB), 1999. WorldS-
can: The Core Version. CPB Netherlands Bureau for Economic Policy
Analysis, The Hague.
Netherlands Bureau for Economic Policy Analysis (CPB), 2003. Four
Futures of Europe. CPB Netherlands Bureau for Economic Policy
Analysis, The Hague.
Davis, G., 2002. Scenarios as a tool for the 21st century. Probing
the Future.In: Conference Group External Affairs, vol. 19. pp. 7–
20.
European Environment Agency (EEA), 2003. Europe’s environment: the
third assessment. Environmental Assessment Report, 10. European
Environment Agency, Copenhagen.
European Commission, 1999. Europe’s Agenda 2000. Strengthening and
Widening the European Union. European Commission, Brussels.
European Commission, 2004. The Agriblue Blueprint. Sustainable Terri-
torial Development of the Rural Areas of Europe. European Commis-
sion, Brussels.
European Commission, 2005. Agriculture, Environment, Rural Develop-
ment: Facts and Figures—A Challenge For Agriculture. European
Commission, Brussels.
European Commission Internet document, 2005. Lisbon Strategy. Putting
Rural Development to Work for Jobs and Growth. European Commis-
sion, Brussels.
European Communities, 2003. Rural development in the European Union.
Fact Sheet. European Communities, Luxembourg.
European Communities, 2004. New perspectives for EU rural development.
Fact sheet. European Communities, Luxembourg.
EURuralis, 2004. EURuralis 1.0 A Scenario Study on Europe’s Rural Areas
to Support Policy Discussion. Alterra, Wageningen (CD-ROM).
Ewert, F., Rounsevell, M.D.A., Reginster, I., Metzger, M., Leemans, R.,
2005. Future scenarios of European agricultural land use. I: Estimating
changes in crop productivity. Agric. Ecosyst. Environ. 107, 101–
116.
Food and Agriculture Organization (FAO), 2003. World Agriculture:
Towards 2015/2030. An FAO Perspective. FAO, Rome.
Gallopın, G., Raskin, P., 2002. Global Sustainability: Bending the Curve.
Routledge, London.
Hertel, T.W., 1997. Global Trade Analysis: Modeling and Applications.
Cambridge University Press, Cambridge.
Hilderink, H. 2000. World population in transition: an integrated regional
modeling framework. Ph.D. Thesis. Thela thesis, Amsterdam.
IMAGE-Team, 2001. The IMAGE 2.2 Implementation of the SRES Sce-
narios. A Comprehensive Analysis of Emissions, Climate Change and
Impacts In the 21st Century. RIVM CD-ROM publication, 481508018.
National Institute for Public Health and the Environment, Bilthoven.
Intergovernmental Panel on Climate Change (IPCC), 2000. Special Report
on Emissions Scenarios. Cambridge University Press, Cambridge.
Kemp-Benedict, E., Heaps, C., Raskin, P., 2002. Global Scenario Group
Futures. Technical notes (revised and expanded). SEI PoleStar Series
Report, 9. Stockholm Environment Institute-Boston, Boston.
Klijn, J.A., Vullings, L.A.E., van den Berg, M., van Meijl, H., van
Lammeren, R., van Rheenen, T., Veldkamp, T., Verburg, P., Westhoek,
H., Eickhout, B., 2005. The EURURALIS study: technical document.
Alterra-rapport, 1196. Alterra, Wageningen.
Knickel, K., Kok, K., 2003. Future land use in Europe: trends, challenges
and policy (FLU-E). EFEIA Scoping Paper. EFEIA, Frankfurt.
Mc Glade, J., 2004. Changing land use in Europe—getting the picture. In:
Speech to EU conference ’Changing land use in Europe’, Kasteel
Groeneveld.
Organisation for Economic Co-Operation and Development (OECD), 2001.
Environmental Outlook. OECD, Paris.
Potting, J., Bakkes, J., 2004. The GEO-3 scenarios 2002–2032: quantifica-
tion and analysis of environmental impacts. RIVM Report, 402001022,
UNEP/DEWA and RIVM, Bilthoven.
Raskin, P., Banuri, T., Gallopın, G., Gutman, P.H.A., Kates, R., Swart, R.,
2002. Great transition: the promise and lure of the times ahead. PoleStar
Series Report, 10. Stockholm Environment Institute-Boston, Boston.
Raskin, P., Chadwick, M., Jackson, T., Leach, G., 1996. The sustainability
transition: beyond conventional development. PoleStar Series Report, 1.
Stockholm Environment Institute, Stockholm.
G. Busch / Agriculture, Ecosystems and Environment 114 (2006) 121–140140
Raskin, P., Gallopın, G., Gutman, P., Hammond, A., Swart, R., 1998.
Bending the curve: towards global sustainability. PoleStar Series
Report, 8. Stockholm Environment Institute-Boston, Boston.
Rotmans, J., van Asselt, M., Inastasi, C., Rothman, D., Greeuw, S., van Bers,
C., 2001. Integrated visions for a sustainable Europe. Final Report to
EC Environment and Climate Research Programme (1994–1998):
Research Theme 4: Human Dimensions of Environmental Change.
ICIS, Maastricht.
Rounsevell, M.D.A., Ewert, F., Reginster, I.L.R., Carter, T.R., 2005. Future
scenarios of European agricultural land use. II: Estimating changes in
land use and regional allocation. Agric. Ecosyst. Environ. 107, 117–
135.
Rounsevell, M.D.A., Reginster, I., Araujo, M.B., Carter, T.R., Dendoncker,
N., Ewert, F., House, J.I., Kankaanpaa, S., Leemans, R., Metzger, M.J.,
Schmit, C., Smith, P., Tuck, G., 2006. A coherent set of future land
use change scenarios for Europe. Agric. Ecosyst. Environ. 114, 57–
68.
Schroter, D., 2004. ATEAM Final Report 2004 Section 5 and 6 and Annex 1
to 6. Reporting period: 01.01.2001-30.06.2004 Contract n8EVK2-2000-
00075. Potsdam Institute for Climate Impact Research, Potsdam.
Sere, C., Steinfeld, H., 1996. World Livestock Production Systems. Current
status, Issues And Trends, vol. 127. Food and Agriculture Organization
of the United Nations, Rome.
Strengers, B.J., 2001. The Agricultural Economy Model in IMAGE 2.2.
RIVM report, 481508015. RIVM, Bilthoven.
United Nations Environment Programme (UNEP), 2002. Global Environ-
ment Outlook. Past, Present And Future Perspectives, vol. 3. Earthscan,
London.
United Nations, 2000. World Population Prospects: The 1998 Revision.
United Nations Department of Economic and Social Affairs, Population
Division, NY.
van Meijl, H., van Rheenen, T., Tabeau, A., Eickhout, B., 2006. The impact
of different policy environments on agricultural land use in Europe.
Agric. Ecosyst. Environ. 114, 21–38.
Verburg, P.H., Schulp, C.J.E., Witte, N., Veldkamp, A., 2006. Downscaling
of land use change scenarios to assess the dynamics of European
landscapes. Agric. Ecosyst. Environ. 114, 39–56.
Wetenschappelijke Raad voor het Regeringsbeleid, 1992. Ground for
Choices: Four Perspectives for the Rural Areas in the European Com-
munity. Sdu Uitgevers, The Hague.