Linking pan-European land cover change to pressures on biodiversity - Biopress final report 1st January 2003 - 31st December 2005, sections 1 to 4

Gerard, F.; Petit, S.; Smith, G.; Thomson, A.; Brown, N.; Manchester, S.; Wadsworth, R.; Bugar, G.; Halada, L.; Bezak, P.; Boltiziar, M.; De badts, E.; Halabuk, A.; Mojse, M.; Petrovic, F.; Gregor, M.; Hazeu, G.; Mucher, C.A.; Wachowicz, M.; Huitu, H.; Tuominen, S.; Kohler, R.; Olschofsky, K.; Ziese, H.; Kolar, J.; Sustera, J.; Luque, S.; Pino, J.; Pons, X.; Roda, F.; Roscher, M.; Feranec, J.. 2005 Linking pan-European land cover change to pressures on biodiversity - Biopress final report 1st January 2003 - 31st December 2005, sections 1 to 4. NERC/Centre for Ecology & Hydrology, 70pp. (CEH Project Number: C02129, EC-FPV Contract Ref: ENV-CT-2002-00178)

Before downloading, please read NORA policies.

Download (3MB) | Preview


BIOPRESS – Linking pan-European land cover change to pressures on biodiversity – is a 3 year EC-FPV project funded in the framework of the GMES ‘Global Monitoring for Environment and Security’ initiative ( It was the only GMES project under the priority theme "Land cover change in Europe”. BIOPRESS’s main goal was to provide the EU-user community with quantitative information on how changes in land cover and land use has affected the environment and biodiversity in Europe. The project aimed at producing consistent and coherent sets of historical (1950 – 1990 – 2000) land cover change information in and around circa 75 Natura2000 sites located from the boreal to the Mediterranean, and from the Atlantic to the continental regions of Europe. These land cover change statistics would be converted into quantitative measures of pressures on biodiversity through the integration of socio-economic indicators. The impact of the land cover changes on biodiversity would also be assessed. The change statistics were produced by means of two parallel activities, the backdating of CORINE land cover 1990 of circa 75 windows (30km x30km) with aerial photography of the 1950’ies and, the interpretation of aerial photography from 1950, 1990 and 2000 for circa 50 transects (2km x 15km). The windows were interpreted to identify the CORINE level 3 land cover and use classes to a minimum mapping unit of 25 ha. The transects, at the other hand, were interpreted to a minimum mapping unit of 0.5 ha. Scientific achievements: Data access The BIOPRESS team established an operational online access point for metadata and data of relevant European datasets. The European data policy appears to be the major obstacle for easy access to European datasets even in case of projects that are financed by the European Commission. The INSPIRE initiative as well as the GMES framework could benefit from the experiences made in the BIOPRESS project in order to streamline access to European wide data relevant for environmental monitoring. Land Cover change The methodological development for production of land cover change matrices was completed successfully ensuring the BIOPRESS team had the appropriate tools (list of 30km x 30km window sites, list of 2km x 15km transect sites, interpretation manual, quality assurance protocol and meta database designed to follow progress) and material (aerial photography) to successfully carry out the photo to photo (1950 – 1990 – 2000) interpretation of transects and CORINE Land Cover 1990 backdating (1950-1990). Several of the tools, in particular the interpretation manuals, have the potential of being adopted by GMES services and future EU projects (The GEOLAND and GSELAND projects were given copies of the manuals on request). A total of 57 transects and 73 windows were interpreted. The results were stored in a database. The database will be made available to the wider research community in 2007. The total extent of land cover changes that have occurred within all BIOPRESS windows account only to 9,62 % of the total measured area. In other words, 90,38% of the measured area within the BIOPRESS windows have shown no change of land cover at all. Overall the most important land cover conversions based on CORINE level 2 nomenclature can be summarised as one of the following: • FROM shrub and/or herbaceous vegetation association TO forests, and its inverse conversion, FROM forest TO shrub and/or herbaceous vegetation association • FROM heterogeneous agricultural areas TO urban fabric, as well as TO forest • FROM arable land TO industrial, commercial, and transport units. Because the focus was on biodiversity and historical land cover changes, it was clear from the start that Europe had to be sampled. Bias was introduced in the BIOPRESS samples by (1) relying on an expert to select a superset of samples including Natura 2000 sites and (2) the availability of aerial photography. The project’s resources limited the total number of samples acquired. As a result some bio-geographical regions were under represented in the sample (Boreal and Mediterranean) whilst other regions were over represented (the Alpine, Atlantic, and Continental). So the development of an appropriate extrapolation approach was seen as a challenge from the beginning of the project. The key was to produce information which is useable in the data integration and which is meaningful, and reliable enough for use by our key stakeholder, the EEA. An extensive sensitivity analysis and the development of minimum land cover accordance maps have provided an excellent insight in the acquired land cover change data with respect to samples’ representativeness of biogeographical areas and land cover. Quality assurance and error propagation The following problems were identified as the main sources of possible mistakes and lack of correspondence in windows: • Ambiguity of CLC classes delineation. • Quality of B&W AP. • Availability of ancillary data. • Separation of CLC classes in B&W AP (E.g. burnt areas). • Diversity within class definition. • Occurrence of polygon less than 0.5 ha. • Amalgamation of objects less than 0.5 ha. • Real changes omitted. • Identification of questionable changes. • Identification of point and linear features, questionable, ambiguity and unknown relevance The quality of the input data was comparable for all transects, indicating that the comparability of results between partners and transects was unlikely to have been influenced by the quality of the input data. The date of the aerial photos (1950, 1990 or 2000) proved to have no influence on the thematic consistency of the interpretations whereas the level of thematic detail did have a high impact. The geometric accuracy was more difficult to evaluated, still we found that the controllers identified more spatial structures than the local interpreters. The quality of the interpretation depends on the land cover characteristics of the individual transects or windows. An error model was developed describing each step of the land cover change production chain. We found that most error sources which reduced the interpretation quality such as image quality, unclear class definition and confusion caused by lax use of land cover and land use attributes in the definitions, can be almost completely reduced by using modern data sources and adjusting the interpretation methodology. However, knowledge and experience of the interpreter play an important role in manual visual interpretation of remotely sensed data. Land cover change and pressures A land cover change - pressure association matrix was developed. This matrix enabled the grouping of types of land cover changes related to one of the six pressures under consideration in BIOPRESS: Urbanisation, intensification, afforestation, deforestation, abandonment and drainage. This cross-tabulation matrix is a fundamental starting point in the analysis of land cover change, because it provides a national-scale assessment of not only the losses or gains in the area of specific land categories but what these changes represent in terms of types of pressures However, additional research is needed to analyse this matrix according to its various components in order to gain more insight into the potential processes that determine a pattern of land cover change. It was impossible to derive a simple and practical list of indicators that would consistently explain the pressures on biodiversity. It was clear that there were multiple potential indicators, and the best indicators have not simply appeared out from the extensive data and information that already existed. In essence, the search for a coherent set of pressure indicators was a frustrating and time-consuming activity. It is clear from our effort that a single set of indicators would never explain the whole dynamics of anthropogenic pressures on biodiversity, and would go only some way towards meeting the needs for understanding the observed land cover changes. The main research challenge faced was to define a pattern-process model of land cover dynamics in space and time in order to combine the local level measurement of the land cover changes (e.g. BIOPRESS windows) and the socio-economic indicators of a larger region (e.g. the countries). The proposed multi-representation model is based on the degree of variability in the behaviour of generalised statistics and their dependency of the spatial generalisation of the variable values at different spatial scales. A systematic analysis of spatial coincidence between land cover types in CLC1990 and Annex1 habitat types recorded in Natura 2000 sites was carried out, translated into the EUNIS habitat classification and summarised per Biogeographical region. The work showed that a significant improvement could be made by adopting a regional approach providing neater and more specific links between CLC classes and habitats than what has been available so far. It also identifies what the limitations are in attributing habitat types to CLC classes. We found that the BIOPRESS land cover change product was suitable for quantifying some pressures on biodiversity but quite insufficient for the interpretation of land cover change related to other pressures: • BIOPRESS contributed very positively to the quantification of urbanization across Europe between 1950 and 1990/2000. • BIOPRESS land cover product made a useful contribution to the quantification of afforestation and deforestation across Europe between 1950 and 1990 but that these pressures could be better understood if (i) we had more points in time, closer together and (ii) more information on the condition of forest was derived from remote sensing and/or ancillary data was used to evaluate the ecological value of forested land. • BIOPRESS will have underestimated the extent to which the pressure land abandonment is threatening biodiversity in Europe, in comparison to other existing assessments (e.g. MIRABEL but also national scale statements). However, it would be possible to increase the accuracy and the generic value of the BIOPRESS estimates by (i) broadening the definition of land abandonment i.e. modifying the pressure matrix, so that it matches what is meant in other assessments and (ii) by increasing the number of points in time. • BIOPRESS was probably the first project to provide quantitative estimates about the shift from small scale to more large scale agriculture for such a large sample area across Europe and in this respect, this is a very important contribution to understanding changes in European biodiversity. However it is important to keep in mind that what has been quantified within BIOPRESS was only a small part of what is usually understood by farming intensification in biodiversity assessments. This means that, as was the case for land abandonment, BIOPRESS results will greatly underestimate the pressure farming intensification, compared to other assessments. The main conclusion is that remote sensing products such as the BIOPRESS land cover change product can provide very helpful information in the field of biodiversity assessment. There is potential for improving this information, e.g. by adding time steps in the monitoring or using external data to help in the interpretation of land cover change. However, our work also shows that there are clear limitations in this contribution and that remote sensing will only provide part of the information. One important recommendation that would lead to improve facilities for large scale biodiversity monitoring would be the integration of remote sensing products with in situ information. This recommendation forms the basis of a position paper produced jointly by BIOHAB (FP5 funded Concerted Action) and BIOPRESS. Socio-economic relevance and policy implications: The project supported the needs of DG-Environment and EEA in helping to implement and assess European policy on nature and biodiversity and contribute to the objective of enhancing the quality of the environment by helping to understand pressures on biodiversity arising from land cover change in the member states and accession countries. The state of the environment is perceived as an important indicator of a high quality of life by a majority of European citizens. The European public increasingly expresses the wish to be informed by policy on the perceived threats to biodiversity. BIOPRESS supported the development of a European capacity for monitoring the state of the environment (GMES) to meet these information needs. Conclusions: BIOPRESS was one of the first wave of thematic projects which were funded through the GMES initiative. As a result its main objective was to produce information at European level which in the case of BIOPRESS was information on historical land cover change for the purpose of assessing past pressures on habitats and their associated biodiversity. A large part of the project’s resources were used to deliver the land cover change database successfully and the outcome has not only been the delivery of data but also a set of tools for future European wide land cover monitoring. The real challenge was when trying to establish a link between land cover change and pressures on biodiversity. The development of the land cover change - pressure association matrix as a first step enabled the grouping of types of land cover changes related to one of the six pressures under consideration in BIOPRESS. This matrix has the potential to enhance the similar ‘Land Cover Flow’ matrix developed by the EEA as part of the EEA Land Accounting System. In theory the idea of integrating socio-economic data with land cover change data made sense but in practice the team struggled with the wide variety of data types, spatial and temporal resolutions. To assess the consequences of the observed Land cover changes on habitats and their biodiversity, BIOPRESS impact tables were developed using the same conceptual approach as that establish for the DPSIR assessment MIRABEL. The overall agreement between MIRABEL and the BIOPRESS tables, which unlike MIRABLE provided quantitative estimates for a selected sample of land in each region, was an important result. This part of the work concluded that a land cover change product such as that produced by BIOPRESS was suitable for quantifying some pressures on biodiversity but quite insufficient for the interpretation of land cover change related to other pressures. The error propagation, quality assessment and data search exercises highlighted the importance of the availability of good quality, affordable data (e.g. aerial photography, digital elevation data, social and economic indicators) which for long term monitoring should be continuously collected in a consistent manner.

Item Type: Publication - Report (UNSPECIFIED)
Programmes: CEH Programmes pre-2009 publications > Biodiversity
CEH Programmes pre-2009 publications > Biodiversity > CC01A Detection and Attribution of Change in UK and European Ecosystems
CEH Sections: Harding (to July 2011)
ISSN: 0309-1333
Funders/Sponsors: European Commission
Additional Information. Not used in RCUK Gateway to Research.: Contact France Gerard (
Additional Keywords: land cover change, remote sensing, historical, Europe, aerial photographs, BIOPRESS, NATURA 2000, impact, biodiversity, GMES
NORA Subject Terms: Ecology and Environment
Related URLs:
Date made live: 19 Jun 2013 13:18 +0 (UTC)

Actions (login required)

View Item View Item

Document Downloads

Downloads for past 30 days

Downloads per month over past year

More statistics for this item...