Conference on Changing Land Use in Europe - Getting the Picture

Kasteel Groeneveld, 10 November 2004

Conference on Changing Land Use in Europe - Getting the Picture

Professor Jacqueline McGlade
Executive Director, European Environment Agency

Speech to EU Conference 'Changing Land Use in Europe' Kasteel Groeneveld, the Netherlands, 9-12 November 2004

Chairman, Ladies and Gentlemen

First of all I would like to thank you for inviting me to this event and congratulate you on the initiative to gather the European directors for rural affairs in what will hopefully be the start-up of a formal network. The EU Rural Development Regulation is currently being reviewed, which makes this initiative very timely. Exchanging views and experiences from all over Europe is the key to making the rural development policy more coherent and effective.

Europe has been shaped by centuries of traditional farming practices, resulting in a wealth of different landscapes with high cultural and natural value. Landscapes are the physical expression of the way we use natural resources and meet our human needs. In a way, the landscape tells us who we are. That is probably why there is such a keen interest in preserving the variety of European landscapes and supporting the rural economies that have created them. From my perspective rural development is ultimately about reinforcing regional identity.

Which brings me to the subject of today's meeting: land use change and its policy implications. This is of great interest, not only because land use affects the environment in many ways, but also because significant policy interventions are made which influence it. Spatial analysis allows us to develop an integrated view on the effectiveness of Europe's policy efforts and is extremely helpful in linking sectoral and thematic issues.

Before I show you some recent results from relevant work undertaken by the European Environment Agency in this field, I would like to begin by examining some of the ways we perceive patterns. Studying land use change basically starts with pattern recognition. People are generally very good at that, and I am sure this audience is no exception. But first let me take you through a small exercise to test that.

Please take a close look at the following sentence. FEATURE FILMS ARE THE RESULT OF YEARS OF SCIENTIFIC STUDY COMBINED WITH THE EXPERIENCE OF YEARS. Don't worry if it does not make immediate sense. It's not the message that we are interested in.

The only thing I will ask you to do is to count the number of F's in the sentence. Just read the sentence carefully and count the number of F's in the next 30 seconds? Who saw one F? Please raise your hands...Who saw two?.... Who saw three? .....Who saw four?...... Who saw five?........Who saw six?.....Who saw more than six?

Actually, there were six! The interesting thing is that we are so familiar with the patterns, that we do not record the details. Not even if we are explicitly asked to focus on them! (For those who were wrong: the people who see 6 F's tend to be dyslexic....).

Landscape change is also a slow process in which it is often the small detail that is changing, and which may go unnoticed. But ultimately the larger picture will change dramatically. Let me demonstrate this by looking at some striking patterns in the landscape.

The first image shows part of the border between Canada in the north and the USA to the south. The area north of the border consists of rangeland with extensive cattle grazing. South of the border there is intensive arable farming. The border between the two countries shows on the satellite image as a ruler-straight line, dividing the different agricultural regimes.

Physical conditions do not play any role here. The background differences reside in different agricultural policies and farming traditions, underlining the powerful impacts of policy interventions.

Another striking example can be seen in the Netherlands, in an area not far from here. Both pictures were taken in the Flevopolder, an area reclaimed from the sea two decades ago. Both sites are only a few hundred metres apart.

To the left we see the polder as it was intended: optimal spatial planning to accommodate intensive and highly efficient farming. However, it contains very low biodiversity.

To the right we see the result of an unexpected development. By the time this part of the polder was about to be taken into agricultural production, an extensive and beautiful wetland had formed. People realised the aesthetic and ecological value of it and the area was left untouched. Now the Oostvaardersplassen area is one of the major wetlands in Europe with a spectacular water bird fauna, including spoonbills, the emblematic species for the Netherlands under the Birds Directive, avocets and little egrets.

This shows the potential of nature itself and the need for some 'out-of-the-box' thinking. Land is a scarce resource and keen spatial planning (in this case resulting in spatial segregation of functions) can be very effective in creating opportunities for innovative land use that will optimise conditions for agriculture and nature alike.

In the next example, the locations are much further apart.

The left-hand picture shows a landscape in Poland with extensive farming. The plots are very small, crops are varied and extensive grazing is practised. The result is a visually attractive landscape with high biodiversity. But the economic returns to farmers are presently very low.
The right-hand image is from the highly rationalised intensive agriculture in the Netherlands. Plots are big, easily accessible, and intensively used. High yields, therefore, but with a low biodiversity. Without an improved economic perspective for the extensive systems, such as in Poland, we can expect agricultural landscapes of the future to look like those in the Netherlands.

Where intensification is not an option, farmers may stop altogether, resulting in land abandonment.

On the left is an image of extensive pastureland in Lithuania, characterised by semi-natural grassland of high botanic quality. The picture on the right shows a similar system after farming ceased and scrub encroachment took place. The result was a large loss of plant species variety.

The next image from Paris illustrates yet another important driver of land-use change: urbanisation. When we think about land-use change and landscapes, we should not only think in terms of traditional rural areas where farming is dominant.

Many of tomorrow's landscapes may have a predominantly urban character. Urban sprawl is likely to continue invading the countryside as agriculture becomes less important and other rural services take over. Already, changes in recreational uses are becoming more important and leading to dramatic land use changes. This should not be seen only as a threat, but also as providing opportunities!

Monitoring and understanding land use change is very relevant for a number of European policies. Land is a scarce resource that needs careful management and consideration in the context of the Lisbon, Sustainable Development Strategy and 6th environment action programme reviews as well as the EU financial perspective 2007-2013. The majority of EU budget expenditure is currently linked to land use (Common Agriculture Policy, Structural and Cohesion funds, Trans-European Network etc.). This will probably continue to be the case.

So what is the European Environment Agency doing to address this issue? We are currently drawing up an image of land use change through a range of projects that tackle the issue from different perspectives. In particular there are three major projects and data products that we believe are of particular relevance to you, the Rural Directors and those involved in spatial planning.

• First is CORINE Land Cover, which is allowing us to generate European land cover maps using ground-validated satellite images. This information will become a significant basis for further analysis by the Agency and policy-makers
• Second is the IRENA project, which builds on these maps by developing agri-environment indicators that allow a detailed and spatially explicit analysis of the influence of agriculture on the environment;
• And thirdly, PRELUDE, a scenario project that explores possible land use developments and their environmental impacts over the next 30 years.

The CORINE Land Cover inventory is based on Landsat ETM+ satellite images. Land cover is mapped in 44 classes with a resolution of 25 ha. Land cover changes are detected at a resolution of 5 ha. The updating frequency is currently every 10 years. On 17 November the Agency will be launching the results of the 2000 inventory and comparisons made with 1990 to show the changes that have occurred over the past decade.

The essence of the Corine Land Cover approach is that subsequent satellite images are compared. The differences that are detected are analysed using additional information, and then mapped. The interpreted maps of the subsequent inventories can then be reconstructed. This is much more accurate than interpreting the satellite images separately and overlaying the interpreted maps, and ensures minimal noise in the data.

It is of course a very labour intensive process; 29 European countries and more than 100 organisations have participated, with a total cost per inventory of about 13 million EURO. Yet despite the cost the analyses are not only insightful but also invaluable in determining the spatial impacts of policies.

The first example shows the mapped differences between two subsequent samples in the Leipzig area (1990 and 2000). There are significant increases over the decade in the number and extent of urbanised areas, new water bodies, afforestation and recultivation after open pit mining.

We can also produce reconstructed maps for both inventories.

The pattern does not seem to differ very much at first sight, but as noted before, the changes on the ground were in fact quite considerable.

CORINE provides unique material for spatial analysis. The next example shows the very different urbanisation patterns in eastern Germany (diffuse urban sprawl) as compared to Poland and Czech Republic (concentration around the city centres).

There is also evidence of reduced farming activity in eastern Germany, Poland and northern Czech Republic. The pattern from the CORINE analysis suggests that the retreat is biggest in eastern Germany, but these are relative figures and the absolute reduction is probably higher in the other two countries.

Pasture provides in general the most valuable agro-ecosystems. The conversion of pasture into arable cover in the same area is likely to signal a significant local loss of biodiversity.

In the IRENA project we go one step further and analyse land use change in a policy context. The IRENA project is the result of excellent cooperation between EEA, DG Agriculture, DG Environment, DG JRC and EUROSTAT. It aims to operationalise a set of 36 agri-environment indicators introduced by the Commission, which have been categorised according to the DPSIR (driving forces, pressures, state, impacts, and responses) framework.

I will now guide you quickly through the IRENA indicators and the DPSIR framework. IRENA provides a unique set of interrelated indicators that enable users to draw up coherent analyses of related environmental issues.

The first category of indicators addresses the drivers: underlying factors that cause changes in environmental impacts of agriculture. The drivers cause changing pressures; leading to a changing state, with certain environmental impacts; giving rise to policy responses.

Thus IRENA is highly relevant for the Cardiff process: integration of environmental concerns into sectoral policy (in this case the CAP).

I will now give an example of the analytical power of the IRENA indicators and the way that land cover data are used.

For this purpose I will focus on high nature value farmland (HNV), which is recognised as a major environmental asset of the European countryside. Its conservation is an explicit goal of the Rural Development Regulation and the Kiev Resolution on Biodiversity.

HNV farmland is generally characterised by extensive farming practices, associated with high species richness and or concentrations of species of particular conservation concern. Typical examples include:

• small-scale extensive agriculture in Romania with meadows and pasture, where the characteristic species is corncrake
• extensively grazed moorland and upland in the UK, where the characteristic species is black grouse
• Spanish dehesa system with overwintering common cranes
• and the Spanish steppic system with the great bustard.

In our analysis three approaches were used to delineate HNV systems, based on land cover from Corine 1990, information about farming systems and distribution patterns of species.

From the Corine data, we can see that HNV farmland areas are primarily found in Spain, Italy, Greece, Northern UK and Scandinavia.

The picture based on farm systems (data from the Farm Accountancy Data Network, FADN), generally confirms the land cover based distribution, but there are some discrepancies, for example in Greece, where relevant farms are not included in FADN.

However, the distribution pattern of important bird species related to farmland is problematic for several reasons: the relation to farmland is not always straightforward - only breeding birds are included and farmland may only be relevant in certain parts of the lifecycle. There is also a bias in the data: species variety generally declines with increasing latitude.

For this reason only the land cover and farm system estimates are being used for further analysis.

HNV farmland is under pressure from intensification and abandonment. The map of farming across Europe indicates a general intensity and some regional trends; for example in important regions in southern Europe (Spain, Italy) both intensification and extensification processes are occurring simultaneously.

As for biodiversity trends, unfortunately no HNV- specific data are available, so the trends we can document are for common species only - rare species and species characteristic for HNV farmland have not been included.

The general picture is very clear, though: biodiversity has declined sharply on farmland as a result of changed farming practices. And some case studies (for example regarding the black grouse in lowland western Europe) indicate that the situation in HNV areas is even worse.

So what are we doing about it? In the biodiversity arena, there are a wide number of conservation initiatives, but also under the CAP, one of the most relevant instruments — the agri-environment schemes that give farmers financial compensation for taking specific environmental measures — is being used. However the rate of uptake varies greatly across Europe. It is particularly low in southern European countries, where in contrast the share of HNV farmland is relatively high, and in lowland western Europe.

Not surprisingly, there appears to be no relationship between agri-environment expenditure and the occurrence of HNV farmland: areas with a high level of HNV farmland, for example Spain and Portugal, get little support. But this is not consistent with the changed rationale of the CAP, with its decoupled payments and increased emphasis on non-trade concerns.

It would therefore seem justified to reconsider the targeting of policy efforts and to pay more attention to those areas with the highest added value in terms of ecological status and resource potential.

But to do this effectively, we will also need to take a number of long-term developments into account. Some drivers of land use change can be affected by policies, others much less so. Some are rather predictable, others are very uncertain. To judge the long term potential of policy interventions, one has to gain insight into major context changes that may happen.

Here is where scenario analysis comes into its own. The EEA has launched the PRELUDE project. This project looks up to 30 years ahead, focusing on the major uncertainties regarding land use change.

The name PRELUDE refers to the 'pilot' character of this project: a try-out of a new method to address future environmental issues and actively influence the policy agenda in an early stage of the policy lifecycle. The name PRELUDE stands for PRospective Environmental analysis of Land Use Development in Europe)

The PRELUDE project is entirely driven by stakeholders. A stakeholder panel with 22 members from all over Europe has drawn up qualitative storylines of how the future might unfold. These storylines are then underpinned by quantitative modelling. Storylines and modelling results are refined in an iterative process.

Two stakeholder meetings have already been held. A third and final one is foreseen for early 2005, followed up by strategic reach-out workshops for a wider audience called 'PRELUDE2Action'. No concrete model results can be presented at this stage, but the concept is as follows:

The different stories are based on different assumptions about drivers of land use change. First the overall changes are calculated, based on the driver assumptions. Then the changes are made spatially explicit using a land use model with allocation rules. Finally the environmental consequences of the changes are analysed.

To stress my point that context changes have to be taken into account, I would like to draw your attention to climate change.

The current projections indicate major changes in crop growing potential and flooding risks. The 30 year perspective shows that climate change will certainly influence agriculture and settlement patterns. We will therefore need to climate proof our current policies to adapt to these changes.

Another large contextual change is Europe's rapidly ageing population. No European country has sufficient net reproduction to maintain its population over one generation of time!

It is also highly unlikely that immigration will be able to compensate for this. Large parts of Europe may effectively depopulate, notably in areas where high nature value farmland predominates - already we can see that the proportion of elderly in these areas is very high. This will have implications for rural development policy!

So to conclude my speech, I would like to give you the following considerations:

• Data and tools are now available for detailed analysis of current land cover/land use trends. These will be very helpful for integrated assessments and policy analyses.
• The absolute land use changes that we can detect may seem small, but in the long term land use patterns may change drastically.
• These changes have major implications for the environment and for rural development.
• Policies in general, and rural development measures in particular, must be judged against long term context changes relating not only to the economy but also to climate change and demography.

I hope that these insights from the Agency's work have helped you to recognise some of the issues that you will be dealing with and I wish you a very fruitful and pleasant meeting.

Thank you for your attention.