Landscape fragmentation pressure and trends in Europe

Assessment versions

Published (reviewed and quality assured)

• 13 Dec 2019, 03:52 PM
  - Landscape fragmentation pressure and trends in Europe

Rationale

Justification for indicator selection

Landscape fragmentation, as described in this indicator, is understood to be the physical disintegration of continuous ecosystems, habitats or landscape units, excluding freshwater ecosystems. Such disintegration into smaller units, or patches, is most often caused by urban or transport network expansion. An important consequence of fragmentation is the increased isolation of newly formed fragments of ecosystems. Breaking structural connections results in decreased resilience and a decrease in the ability of habitats to provide various ecosystem services. Furthermore, it prevents access to resources for wildlife, reduces habitat area and quality, and may isolate some wildlife populations, resulting in smaller and more vulnerable fractions. Reducing habitat degradation and fragmentation may ensure that those habitats that remain are more capable of supporting biodiversity. Finally, yet importantly, fragmentation not only directly affects fauna and flora, but also indirectly influences human communities, agriculture, recreation and overall quality of life. Fragmentation decreases landscape quality and changes the visual perception of landscapes, thus decreasing the attractiveness of landscapes for recreational activities.

Scientific references

• Jaeger, J. A.G. (2000): Landscape division, splitting index, and effective mesh size: New measures of landscape fragmentation. Landscape ecology 15(2), pp 115-130.
•  Moser, B., Jaeger, J.A.G., Tasser, E., Eiselt, B., Tappeiner, U. (2007): Modification of the effective mesh size for measuring landscape fragmentation to solve the boundary problem. Landscape Ecology 22,pp 447–459.
•  Jaeger, J.A.G , Bertiller, R., Sccwick, Ch., Muller, K., Steinmeier, CH, Evald, K.C., Ghazou, J.: Implementing Landscape Fragmentation as an Indicator in the Swiss Monitoring System of Suistanable Development (MONET). Journal of Environmental Management
• Jaeger, J.A. G., Mandrinan, L.F., Soukup, T., Schwick, Ch., Kienast, F. (2011): Landscape fragmentation in Europe. Joint EEA-FOEN report.
• EEA / FOEN, 2016. Urban sprawl in Europe. Joint EEA-FOEN report No 11/2016. European Environment Agency and Federal Office for the Environment.
• EEA / FOEN, 2011. Landscape fragmentation in Europe. Joint EEA-FOEN report. European Environment Agency and Federal Office for the Environment.
• EEA, 2011. Increasing fragmentation of landscape threatens European wildlife. EEA News.
• Eurostat, 2011. GEOSTAT 1A Representing Census data in a European population grid.
• Plan Blue, 2016. Tourism. Economic activities and sustainable development. In: Building the Mediterranean future together. Plan Bleu notes.
• WTTC, 2015. Economic impact of Travel and Tourism in the Mediterranean.
• Barrington-Leigh, C., Millard-Ball, A., 2017: The world’s user-generated road map is more than 80% complete. https://doi.org/10.1371/journal.pone.0180698
• EARSC Position Paper on Usage of Open Street Map versus National Data for CORINE Land Cover plus (CLC+) – “CLC Backbone"

Indicator definition

This indicator measures landscape fragmentation due to transport infrastructure and sealed areas. Unlike the previous indicator on fragmentation status, this updated version uses the TeleAtlas® Multinet data set to ensure the statistical comparability of the time series. While the Open Street Map data set is a valuable source of the street network available for the general public, there are still inconsistencies in this data set for some regions of Europe, which render it secondary to the TeleAtlas data set.

As in the previous version, this indicator is based on the effective mesh size method (Jaeger, 2000). For some species, the effective mesh size (meff) can be interpreted as the area that is accessible when beginning to move from a randomly chosen point inside a landscape without encountering anthropogenic barriers such as transport routes or built-up areas. However, it should be stressed that for many species that can fly, or are effective dispersers in others ways, man-made structures may not act as barriers. The combination of all barriers in a landscape is referred to as the fragmentation geometry (FG) hereafter.

The meff value expresses the probability that any two points chosen randomly in an area are connected. Hence, meff is a measure of landscape connectivity, i.e. the degree to which movements between different parts of the landscape are possible. The larger the meff, the more connected the landscape. The indicator addresses the structural connectivity of the landscape and does not tackle functional, species-specific connectivity.

The effective mesh density (seff) is a measure of landscape fragmentation, i.e. the degree to which movement between different parts of the landscape is interrupted by fragmentation geometry. It gives the effective number of meshes (or landscape patches) per 1 000 km², in other words the density of the meshes. The seff value is 1 000 km²/meff, hence the number of meshes per 1 000 km². The more barriers fragmenting the landscape, the higher the effective mesh density.

The values of meff and seff are reported within the cells of a 1 km² regular grid.

The value of meff is area-proportionally additive, hence it characterises the fragmentation of any region considered, independently of its size, and thus can be calculated for a combination of two or more regions. It has several advantages over other metrics:

• It addresses the entire landscape matrix instead of addressing individual patches.
• It is independent of the size of the reporting unit and its values can be compared among reporting units of differing sizes.
• It is suitable for comparing the fragmentation of regions with differing total areas and with differing proportions occupied by housing, industry and transportation structures.
• Its reliability has been confirmed on the basis of suitability criteria through a systematic comparison with other quantitative measures. The suitability of other metrics is limited, as they only partially meet the following criteria:
• intuitive interpretation;
• mathematical simplicity;
• modest data requirements;
• low sensitivity to small patches;
• detection of structural differences;
• mathematical homogeneity (i.e. intensive or extensive).

Units

Values of meff are positive real numbers, where 0 stands for grid cells completely covered by urban areas and infrastructure (i.e. the landscape is covered by impermeable surfaces). The lower threshold for meff is 0.000001 km² (= 1 m²), and smaller values are rounded to this value. The highest possible value of meff is limited by the area of the landscape patches as well as by the area of the fragmentation geometry affecting the landscape patches. A landscape patch is defined as a continuous area with the barriers of the fragmentation geometry as boundaries. Hence, the largest meff value will be assigned to the largest continuous landscape patch with the smallest area taken up by the fragmentation geometry (see illustration in "Methodology for indicator calculation" section).

The seff values are positive real numbers. If meff = 0.000001 km², then seff = 1 000 000 000 meshes per 1 000 km². For grid cells completely covered by built-up areas and infrastructure (i.e. where meff = 0 km²), the seff value is set to -2, i.e. -2 represents positive infinity.

For convenience and practical considerations, meff values of < 0.01 km² (= 10 000 m²) are rounded to 0, as these values are too small to be measurable without noise on a European scale. As a consequence, the largest reported seff value is 100 000 (= 1 000 km²/0.01 km²) meshes per 1 000 km².

This indicator presents seff, rather than meff, values because these are more intuitive to understand as indications of fragmentation. For the assessment, seff values were grouped into five fragmentation classes (very low, low, medium, high and very high) by performing the following steps:

(1)                selecting 95 % of the seff value range (ignoring the upper and lower 5th percentiles);

(2)                running geometric interval classification;

(3)                rounding threshold values for straightforward comparisons and change detection.

The thresholds for the fragmentation classes are outlined below.

Policy context and targets

Context description

Priority objective 1, paragraph 23, of the Seventh Environment Action Programme (7th EAP) explicitly lists fragmentation as one of the key elements necessary to protect, conserve and enhance the Union’s natural capital: 'The degradation, fragmentation and unsustainable use of land in the Union is jeopardising the provision of several key ecosystem services, threatening biodiversity and increasing Europe’s vulnerability to climate change and natural disasters. It is also exacerbating soil degradation and desertification.'

Furthermore, priority objective 7 ('To improve environmental integration and policy coherence'), paragraph 87, offers ample space for fragmentation to play a role in the more effective integration of environmental and climate-related considerations into other policies: 'Incorporation of the green infrastructure can also help to overcome the fragmentation of habitats, preserve and restore ecological connectivity, enhance ecosystem resilience and thereby ensure the continued provision of ecosystem services, including carbon sequestration, and climate adaptation, while providing healthier environments and recreational spaces for people to enjoy.'

The EU 2020 biodiversity strategy, specifically target 2, indirectly addresses the fragmentation of ecosystems and habitats, as it requires that 'by 2020, ecosystems and their services are maintained and enhanced by establishing green infrastructure and restoring at least 15 % of degraded ecosystems'.

Reducing fragmentation will also contribute to achieving all other targets of the EU biodiversity strategy, such as target 1 concerning the full implementation of the Birds and the Habitats Directives. In particular, paragraph 1 of Article 3 of the Habitats Directive sets up the legal framework for the Natura 2000 network, and paragraph 3 states that 'Where they consider it necessary, Member States shall endeavour to improve the ecological coherence of Natura 2000 by maintaining, and where appropriate developing, features of the landscape which are of major importance for wild fauna and flora, as referred to in Article 10.'

In addition, Article 6.4 of the Habitats Directive stipulates that Member States are to take 'all compensatory measures necessary to ensure that the overall coherence of the Natura 2000 network is protected'. Article 10 of the Habitats Directive and Article 3 of the Birds Directive also include more general connectivity provisions that relate to land use planning and development policies. Work on the fragmentation of ecosystems and habitats will also contribute to achieving targets 3 and 4 of the EU 2020 biodiversity strategy concerning maintaining and enhancing biodiversity in the wider countryside (and the marine environment).

[1] EEA, 2014, Fragmentation: Overview of the knowledge base in the field of habitat and landscape fragmentation.

Targets

None of the existing EU policies sets quantitative targets for reducing and/or measuring the harmful impacts of the fragmentation of ecosystems. The EU 2020 biodiversity strategy, specifically target 2, directly addresses the fragmentation of ecosystems and habitats, as it requires that 'by 2020, ecosystems and their services are maintained and enhanced by establishing green infrastructure and restoring at least 15 % of degraded ecosystems'. 

Combating fragmentation will contribute to achieving all other targets of the EU biodiversity strategy as well, such as target 1 concerning the full implementation of the Birds and the Habitats Directives. In particular, paragraph 1 of Article 3 of the Habitats Directive sets up the legal framework for the Natura 2000 network, whereas paragraph 3 states that 'Where they consider it necessary, Member States shall endeavour to improve the ecological coherence of Natura 2000 by maintaining, and where appropriate developing, features of the landscape which are of major importance for wild fauna and flora, as referred to in Article 10.'

In addition, Article 6.4 stipulates that Member States are to take 'all compensatory measures necessary to ensure that the overall coherence of Natura 2000 is protected'. Article 10 of the Habitats Directive and Article 3 of the Birds Directive also include more general connectivity provisions that relate to land use planning and development policies. Work on the fragmentation of ecosystems and habitats will also contribute to achieving targets 3 and 4 of the EU 2020 biodiversity strategy concerning maintaining and enhancing biodiversity in the wider countryside and the marine environment.

Related policy documents

• Council Directive 92/43/EEC of 21 May 1992
  Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora.
• EU 2020 Biodiversity Strategy
  in the Communication: Our life insurance, our natural capital: an EU biodiversity strategy to 2020 (COM(2011) 244) the European Commission has adopted a new strategy to halt the loss of biodiversity and ecosystem services in the EU by 2020. There are six main targets, and 20 actions to help Europe reach its goal. The six targets cover: - Full implementation of EU nature legislation to protect biodiversity - Better protection for ecosystems, and more use of green infrastructure - More sustainable agriculture and forestry - Better management of fish stocks - Tighter controls on invasive alien species - A bigger EU contribution to averting global biodiversity loss
• European Landscape Convention
  European Landscape Convention
• General Union Environment Action Programme to 2020 - Living well, within the limits of the planet
  7th EAP. ISBN 978-92-79-34724-5 doi:10.2779/66315
• Pan-European Biological and Landscape Diversity Strategy
  Pan-European Biological and Landscape Diversity Strategy

Methodology

Methodology for indicator calculation

The calculation of the effective mesh size (meff) is based on three spatial data sets: (1) the 'landscape' extent, (2) the fragmentation geometry (FG) (landscape elements representing man-made barriers) and (3) reporting units (spatial units for which meff is calculated). The following steps are followed in computing the indicator.

Step 1: landscape extent

The 'landscape' for the calculation of meff is the seamless area of Europe. The input for this step is the Copernicus high resolution layer (HRL) on imperviousness density (IMD) from 2012 [1].

Step 2: fragmentation geometry

Fragmentation geometries are man-made landscape elements, which divide the landscape into unconnected patches. Only anthropogenic elements are considered because the indicator addresses fragmentation of the landscape in urban areas and from transport infrastructure (road and rail).

Step 2.1: fragmentation geometry — built-up areas

Built-up areas are excluded during the 'landscape extent' preparation step. From this layer, a binary mask is created and pixels with IMD values of > 30 % are deleted from the data set. 

Step 2.2: fragmentation geometry — road network

The data set representing the transportation network must meet the following technical requirements:

1. It has to be methodologically stable, so that changes in time represent real changes and not the level of data set completion.
2. The nomenclature/classes of roads must be clearly defined, and consistent over time, to allow different levels of fragmentation detail.
3. It must be topologically correct, i.e. it must not contain discontinuities.
4. It must enable the differentiation of landscape elements that have a major impact on the resulting connectivity or isolation of patches, such as tunnels, overpasses, etc. (where such elements occur, landscape patches may in fact be interconnected and thus the value of fragmentation can be considerably different).
5. It must be based on regularly updated and if possible open source data streams to ensure the sustainability of the indicator.
   
   

The TeleAtlas® Multinet data set^]was used to process the road network fragmenting geometry. The following road/rail classes were included (road class numbering is based on FRC attribute values):

0 — motorway, freeway
1 — major road less important than a motorway
2 — other major road
3 — secondary road
4 — local connecting road

Railroads

The line vectors were buffered according to the road classes to create polygon objects. Buffer sizes were selected according to the road class they represent. Buffering was also applied to prevent small topological inconsistencies in the TeleAtlas data set.

Table 1: Buffers applied to the various TeleAtlas road and rail classes

The result of step 2 is a fragmentation geometry layer that contains landscape patches (i.e. polygons representing the remaining non-fragmented areas) and gaps (no value), in locations where fragmentation geometries were deleted from the landscape.

Methodology for gap filling

The Copernicus HRL is based on satellite imagery classification. As such, there are areas assigned with no IMD values because of cloud coverage (satellite data sets are sometimes not cloud free). These gaps in the data set are filled using the Corine Land Cover (CLC) data set using the corresponding build-up mask derived from CLC classes. These are:

• 1.1. continuous urban fabric, discontinuous urban fabric;
• 1.2. industrial and commercial units, road and rail networks and associated land, port areas and airports;
• 1.3. mineral extraction sites, dump sites and construction sites;
• 1.4.2. sport and leisure facilities (only included if they were completely surrounded by the previous classes);
• 4.2.2. salines.

Methodology references

• Moser, B., Jaeger, J.A.G., Tasser, E., Eiselt, B., Tappeiner, U. (2007). Modification of the effective mesh size for measuring landscape fragmentation to solve the boundary problem. Landscape Ecology 22,pp 447–459. 

Data specifications

EEA data references

• Copernicus Land Monitoring Service - High Resolution Layers - Imperviousness provided by European Union

External data references

• Multinet street map (Dataset URL is not available)

Data sources in latest figures

Uncertainties

Methodology uncertainty

The methodology is without any major uncertainty. Some critique might arise regarding the fragmentation geometries, which were included (or not included) as barriers. This is however not a methodological uncertainty of m_eff and s_eff, but is rather a matter of consciously addressing the spatial detail of the indicator.

Data sets uncertainty

Uncertainty of the Copernicus HRL IMD data set: clouds are contained in the data layer. Corresponding Copernicus CLC data are used for the map filling (see 'Methodology for gap filling' section). Because the spatial resolutions of the HRL IMD and CLC data are different, the spatial detail of the indicator may be influenced for the cloudy area. The metadata layer is part of the indicator data set indicating HRL IMD cloud areas. 

Uncertainty of the Open Street Map (OSM) data set: the maturity, completeness and classification stability of the OSM data set are critical features for monitoring indicator changes and trends. Based on the OSM stability analysis done in 2016 (see link below), these qualities have been confirmed. Nevertheless, as OSM is a collaborative project providing crowd-sourced data under the Open Database Licence, the data set will have to be carefully analysed before any subsequent indicator update.

https://forum.eionet.europa.eu/etc-urban-land-and-soil-systems/library/10.-ap-2018/1.8.1.4-re-analysis-landscape-fragmentation-time-series/osm-stability-analysis-document

Rationale uncertainty

There is no rationale uncertainty.