Indicator Assessment

Species diversity

Indicator Assessment
Prod-ID: IND-9-en
  Also known as: CSI 009
Published 29 Nov 2005 Last modified 11 May 2021
17 min read
This is an old version, kept for reference only.

Go to latest version
This page was archived on 23 Feb 2018 with reason: No more updates will be done

Butterfly and bird species occurring in different habitat types across Europe show population declines of between -2% and -37% since the early 1970s. Similar trends can be observed in the land-cover change for related habitats between 1990 and 2000, especially for heaths and scrubs as well as  mires, bogs and fens, which are specific wetland habitats.


Required information is not filled in: Information about the starting date of the publishing schedule is missing.

Land cover change from 1990 to 2000 expressed as % of the 1990 level, aggregated into EUNIS habitat level 1 categories

Note: The table shows the land cover changes expressed as % of the 1990 level - aggregated into EUNIS habitat level 1 categories.

Data source:

Corine Land Cover, EEA

Temporal coverage for the three data sets

Note: This figure shows the temporal coverage of three different datasets for Habitats, Birds and Butterflies

Data source:

Data provenance info is missing.

The indicator links population trends of species belonging to these two groups (birds and butterflies) to the trends in extent of different habitat types deriving from land cover change analysis 1990-2000.

The assessment is based on 295 butterfly species and 47 birds species linked to 5 different habitat types across several European countries. Results vary among species/habitats groups, but it is striking that both birds and butterflies, linked to different habitat types, show a decline in all habitats examined.

The wetland bird and butterfly species population decline can be explained by direct habitat loss as well as habitat degradation through fragmentation and isolation. Mires, bogs and fens habitats are shown to have the strongest decline in area (-5%) across EU-25 between 1990-2000, a trend based on detecting changes bigger than 25 hectares.

Heaths and scrubs have a particular high diversity of butterfly species, up to at least 92 species of those surveyed. Direct habitat loss (-2%) as well as habitat degradation through fragmentation and isolation also play a role in the very substantial  28% decline observed amongst butterfly species.

The highest number of species assessed, namely 206 butterfly species and 23 birds species, occur in the farmland habitat. These species are typical of open grassy areas such as extensively farmed areas, grasslands, meadows and pastures. The two species groups show very similar trends of decline, namely -28% and -22% respectively. The main pressures related to this decline are loss of extensive farmland with a low or no input of nutrients, herbicides and pesticides, and an increase in agricultural intensification, which leads, among other factors, to loss of marginal habitats and hedgerows and a higher input of fertiliser, herbicides and insecticides.

The area of woodland and forest habitats has increased by 1% since 1990, which in absolute terms is about 500 000 hectares. However, the species linked to the woodland and forest habitats have declined. The 89 butterfly species occurring in this habitat show a decline by -24% compared to a -2% decline of woodland, park and garden birds. Nearly all forests in Europe are managed to some extent and the various management schemes surely have an impact on species diversity. For example, the presence of dead wood and old growth trees are of importance to birds for nesting and feeding and clearing of forest are an important factor for the forest butterflies.


Figure 4 shows there has been a marked decline of European birds since the 1980s. Populations of common and widespread farmland bird species in 2003 are only 71% of their 1980 levels. This is highly significant given the EU objective of halting biodiversity loss by 2010. This decline can be attributed in large part to changes in agricultural practices, including a shift to more intensive practices.
Birds are considered good proxies for biodiversity as they are high in the food chain and so they reflect changes in ecosystems. They also have large European ranges, and are abundant enough to be monitored accurately. The assessment is based on data for 23 common farmland species from 18 European countries.


Supporting information

Indicator definition

Currently the species groups considered are:

  • Birds: farmland, woodland, park and garden birds.

Abundance variation trend over years

Arthropods: butterflies

Distribution variation trend over 20-25 years


No unit


Policy context and targets

Context description

There is a strong need for an indicator to show the status of biodiversity in Europe. Such an indicator should be closely linked to the following policy objectives expressed at both European and global level.

At European level, the Council of the European Union adopted the European Strategy for Sustainable Development in 2001. One of the objectives of the Strategy was "to halt the loss of biodiversity by 2010". In June 2004, the EU Environment Council welcomed the set of biodiversity indicators referred to in the "Message from Malahide" and based on the first set of indicators adopted under the Convention on biological diversity earlier that year. 

Other political instruments in Europe are also focussing on biodiversity. These include the 6th Environmental Action Programme and the European Community Biodiversity Strategy and Action Plan.

On the Pan-European level, the Kiev resolution on Biodiversity was adopted during the fifth ministerial conference on Environment for Europe in 2003. It reinforces the objective to halt the loss of biodiversity at all levels by the year 2010.

At the global level, the Convention on Biological Diversity (CBD), and in particular the Strategic Plan for the Convention commits the Parties to achieve by 2010 a significant reduction of the current rate of biodiversity loss at the global, regional and national level. This target was endorsed in 2002 by two major global environmental meetings; firstly the Ministerial Declaration at COP6 to the CBD and secondly the World Summit on Sustainable Development in its Plan of Implementation (2002).


To halt the loss of biodiversity by 2010

Related policy documents



Methodology for indicator calculation

Selection of farmland birds and woodland, park and garden birds

A selection of 24 common farmland bird species was made but only 23 were assessed by national monitoring schemes.

A selection was made of 24 common woodland bird species characteristic of a range of wooded habitats in Europe. The birds chosen are those characteristic of 'woodland' though many occur in other habitats such as gardens, hedges, scrub and so forth and make use of that habitat too.

These birds all use these specific habitats during their breeding season and also have a large range across Europe. National monitoring coordinators provided with their own assessment - proportion of a species' national population breeding in a given habitat type in four categories (less than 25%, 25 to 50%, 50 to 75%, more than 75%).

 Birds species classified as farmland specialist in the Atlas of European Breeding Birds and in Birds in Europe 
  • Skylark (Alauda arvensis)
  • Linnet (Carduelis cannabina)
  • Goldfinch (Carduelis carduelis)
  • Quail (Coturnix coturnix)
  • Yellowhammer (Emberiza citrinella)
  • Red-backed Shrike (Lanius collurio)
  • Corn Bunting (Miliaria calandra)
  • Tree Sparrow (Passer montanus)
  • Whinchat (Saxicola rubetra)
  • Turtledove (Streptopelia turtur)
  • Starling (Sturnus vulgaris)
  • Whitethroat (Sylvia communis)
  • Lapwing (Vanellus vanellus)
 Bird species classified as farmland specialist by expert judgement 
  • Little Owl (Athene noctua)
  • Greenfinch (Carduelis chloris)
  • Woodpigeon (Columba palumbus)
  • Carrion Crow (Corvus corone)
  • Jackdaw (Corvus monedula)
  • Reed Bunting (Emberiza schoeniclus)
  • Kestrel (Falco tinnunculus)
  • Swallow (Hirundo rustica)
  • Yellow Wagtail (Motacilla flava)
  • Magpie (Pica pica)


 Bird species classified as woodland specialist according to The Atlas of European Breeding Birds and Birds in Europe 
  • Aegithalos caudatus
  • Anthus trivialis
  • Garrulus glandarius
  • Muscicapa striata
  • Parus ater
  • Phoenicurus phoenicurus
  • Phylloscopus collybita
  • Prunella modularis
  • Regulus regulus
  • Sylvia atricapilla
  • Troglodytes troglodytes
  • Turdus philomelos
  • Turdus viscivorus
 Birds classified as woodland specialist according to expert judgement 
  • Accipiter nisus
  • Buteo buteo
  • Dendrocopos major
  • Erithacus rubecula
  • Fringilla coelebs
  • Jynx torquilla,
  • Parus caeruleus
  • Parus major
  • Phylloscopus trochilus
  • Sylvia borin
  • Turdus merula


Calculation for Birds

National trend indices are computed for each species with adequate data for as long as the monitoring scheme has been running within that nation using the program TRIM (Trends & Indices for Monitoring data). TRIM is a program developed by Statistics Netherlands for the analysis of time series counts with missing observations using Poisson regression. Missing counts were estimated from changes in other sites using covariates. The program is used here to produce estimate of yearly indices. The TRIM program then allows combining the country scheme totals as if the raw data were available.

Thereafter, the country results were combined by grouping them into four regions (west, east, north and south) based on geography and similarity in landuse in order for missing data to be imputed by TRIM from comparable countries. Species indices are then combined using a weighting factor to account for the fact that different countries host different numbers of each species. Specifically, the indicators are weighted by the national farmland population size estimates taken from a standard reference (BirdLife International, 2004). The pan-European species indices are then combined into a multi-species, multi-national indicator by combining the indices on a geometric scale.

National bird monitoring data are gathered using a variety of count methods (e.g. point transects/line transects, territory mapping etc), using a variety of sampling strategies (from free choice of plots to stratified random sampling), and individual plot sizes vary within each country (from 1x1km or 2x2km squares, to 2.5 degree grid squares, to irregular polygons).  Despite local variation in approaches, it is well established that the different methods deliver highly comparable and high quality trend information through time. The analysis of trends is highly standardised using the TRIM program.

The method for calculating trends in bird populations, as described above, is based on data collected via the most widely used and well coordinated monitoring network in Europe at present. Moreover the methodology is under continuous development and refinement. The following paragraph gives a brief overview of the strength and weaknesses of the method.

Strengths: The data is based on a strong network of national NGOs that collect ornithological data on a regular basis using standard methods. This effort provides a unique set of data that gives information on farmland bird trends across Europe. Most countries update the data every year. The Pan-European Common Bird Monitoring Project coordinates the overall work. This coordinating function includes collecting and validating the data coming in from all the countries. The data is then analysed and aggregated using standard statistical methods and aggregated to give European trend indices.

Weaknesses: The network suffers from gaps in space and time. Only three countries (Denmark, Sweden and the UK) have collected data since 1980. The rest of the countries joined the network at different years and the time series for the data is therefore very different from country to country. Finland, Greece, Luxembourg, and Portugal have not yet joined this project and data is therefore not available.  The aggregated EU index curve is based on a weighting procedure using estimated of national bird population sizes. For some countries however these estimates are relatively weak. A further improvement of the methodology is under development.

Data are collected nationally on an annual basis during the breeding season using a variety of common bird monitoring schemes but some cover different periods. 

Selection and calculation for Butterflies

Trends in butterfly populations were determined using the Red Data Book on butterflies. This book lists for all 576 species the trends in distribution area from 1975-2000. The assumption made here is that trend in distributions is very conservative compared to trend on populations as a decrease of n % in distribution supposes a decrease of > n % in population.

Distribution, trend and habitat data were collected for each country through a network of more than 50 expert national compilers who each completed a questionnaire. Each expert also gave a list of Corine Land Cover habitats in which the species could be found within their country. For the purpose of this indicator, 17 Corine habitat types have been grouped into 4 major habitat types:

  • heath and scrub
  • grassland
  • forest
  • wetlands.

In order to assess not only the species trends but also the quality of particular habitat types the butterflies have been grouped according to the habitat types. Some species are more specific to a habitat type than other species and are considered a specialist while other species are not dependent on a particular habitat at all and can be considered a generalist species. A specialist species is more often mentioned in one habitat than in the sum of all the others. A generalist species is defined as a species that occurs in at least 10 countries, and on average the species should be mentioned from at least three Corine habitats per country and 1.5 major habitat types.

A total of 325 specialist species and 17 generalist species were identified. The average trend for all species in a major habitat type in a country (for example all wetland specialist species in Belgium) is calculated as follows: The trend reported by each country expert is divided into categories as follows. Each of these categories are converted to a midpoint in order to calculate an average trend for all species of a country.

Extinct   -1
Decrease 75-100% -0.875
Decrease 50-75% -0.625
Decrease 25-50% -0.375
Decrease 15-25% -0.2
Stable   0
Increase 125-200% +0.625
Increase >200% +1


Methodology for gap filling

  • Farmland birds: Trends in adjacent countries were used to fill minor data gaps for some species and countries. See above.
  • Woodland, park and garden birds: Trends in adjacent countries were used to fill minor data gaps for some species and countries. See above. 
  • Butterflies: not relevant

Methodology references

No methodology references available.



Methodology uncertainty

Farmland birds; Woodland, park and garden birds: Species selection is based on expert judgment. It is to be noted that a species selection procedure based on statistical criteria e.g. species occurence within habitats, would provide more relevant results as relation between species trends and biodiversity trends in these habitats would be straightforward and of known confidence.

Butterflies: Distribution data and not  population monitoring data was used.


Data sets uncertainty


  • In data sets
    • Geographical and time coverage on EU level:

One general uncertainty relating to the geographical level and time coverage is that most of the data for this indicator is based on work by NGOs. These NGOs most often work on voluntary basis and are generally dependent on continued funding and resources in order to secure delivery of timely and frequent data updates.


Farmland birds; Woodland, park and garden birds:

On the EU level, data are available for 1980-2002. Data are available for 16 of the EU-25 Member States (not available for  Cyprus, Finland, Greece, Lithuania, Luxembourg, Malta, Portugal, Slovak Republic, Slovenia ) and Norway and Swizerland. The data also show differences among countries in monitoring periods.

Butterflies: past 25 years with no precision and questionnaires filled in late 90's

  • Representativeness of data on national level:

Farmland birds; Woodland, park and garden birds:

The representativity of the data is high on EU level because the selected bird species are widely distributed.

Assumptions that may affect representativeness:

1) A same list of species is considered to be representative of one habitat at European level apart from variation of species habits across his own distribution.

2) We assume that all the monitoring schemes do provide the same quality of information

3) For gap filling, it is assumed that a trend for a country can be derived from neighbour one

4) We assume that the EU trend resulted of 18 countries is informative for all EEA including countries with no monitoring schemes.

Butterflies: Good representativity. The data are from questionnaires filled out by national experts.

  • Comparability:

Farmland birds; Woodland, park and garden birds:

The overall comparability is good. The data collection is based on a Pan-European monitoring scheme using a standardized methodology across countries. For a deeper evaluation of the biodiversity in farmland and forest ecosystems per country a selection of species may be necessary per country, i.e. so different species selected for each country.

Butterflies: good comparability as data collection was fully harmonised.


Rationale uncertainty

One major uncertainty for this indicator in general is, that it is composed of different data types, that may be difficult to aggregate across species groups and across countries to give one overall generic indicator. We are trying to solve this, but for now it is neither possible nor justifiable to show a generic indicator aggregating information from all the data sources presented in this IP. The species and species groups will be presented separately to show their part of the overall picture.

Farmland birds; Woodland, park and garden birds: This indicator is based on a selection of 48 birds that are widespread across Europe and are linked to farmland or to woodland. All countries have the same list of bird species. If we used national selections of birds it may give a stronger link to the particular habitat type.

Butterflies: Only very few countries have butterfly monitoring (UK, NL, BE). The butterfly trends are therefore based on trends in distribution as a proxy to population trends.

Data sources

Other info

DPSIR: State
Typology: Descriptive indicator (Type A - What is happening to the environment and to humans?)
Indicator codes
  • CSI 009
EEA Contact Info


Geographic coverage

Temporal coverage



Filed under:
Filed under: biodiversity, csi
Document Actions