Statistics Explained

Archive:Agri-environmental indicator - landscape state and diversity

This article has been archived. For further information about agri-environment, please see the Agri-environmental_indicators.

This article provides a fact sheet of the European Union (EU) agri-environmental indicator (AEI) landscape - state and diversity. It consists of an overview of data, complemented by all information on definitions, measurement methods and context needed to interpret them correctly. The landscape - state and diversity article is part of a set of similar fact sheets providing a complete picture of the state of the agri-environmental indicators in the EU.

Table 1: Composition of the final index on landscape structure.
Source: Joint Research Centre, European Commission
Figure 1: Structure of the agrarian landscape in Member States according to the degree of dominance of agricultural land use and its diversity in terms of number of crops (%), 1996-2005, EU-27.
Source: Joint Research Centre, European Commission
Map 1: Agrarian landscape physical structure 1996-2005, EU-27.
Source: Joint Research Centre, European Commission
Table 2: Reference scale for hemeroby values.
Source: Joint Research Centre, European Commission
Figure 2: Distribution frequencies of hemeroby classes in agrarian landscape in Member States (%), 1996-2005, EU-27.
Source: Joint Research Centre, European Commission
Map 2: Degree of hemeroby of the agrarian landscape 1996-2005, EU 27.
Source: Joint Research Centre, European Commission
Map 3: Degree of hemeroby of the overall landscape 1996-2005, EU 27.
Source: Joint Research Centre, European Commission
Map 4: Social awareness of the agrarian landscape 1996-2005, EU 27, NUTS 2.
Source: Joint Research Centre, European Commission
Figure 3: Contribution of the three components to the total social awareness indicator (average value by country expressed in relative terms with indices without measurement units), EU-27

The landscape state and diversity indicator describes the main characteristics of the agrarian landscape, in terms of structure of the landscape, cultural influence on the potential natural vegetation due to human activities, and societal awareness of the rural landscape.

In the agri-environmental context the indicator describes:

  • dominance and internal structure of the agrarian landscape in the context of the wider landscape matrix;
  • the hemeroby state, which indicates the degree of influence on land cover and state due to human (agricultural) activities;
  • the interest and perception that society has for the agrarian landscape.

Main statistical findings

Key messages 

Europe has a great variety of agrarian landscapes that reflect differences in biophysical conditions, farm management practices and cultural heritage. In such a context farmers play a crucial role in transforming, managing and maintaining landscapes. Capturing the complexity and multiple functions of European landscapes in one single indicator is not possible, thus three components, each describing a very different aspect of the agrarian landscape are presented instead. These three components are: 

  • the physical structure of the agricultural landscape, intended as land cover and its spatial organisation as a product of land management; 
  • the hemeroby state as a proxy for the influence exerted by farming practices on land cover and state;
  • the societal awareness of the landscape, as the society perceives, assesses and values landscape quality, plans, manages, and uses the landscape for productive or non-productive purposes.

Monitoring these three components will indicate if the trend in landscape structure leads to a higher homogeneity or diversity; how trends in farming practices influence the hemeroby index; if society is becoming more aware of the services the agrarian landscape provides.

Main messages derived from the components are:

  • The structure of the agrarian landscape in terms of dominance of agricultural land use and diversity of crops is organised as follows: 34 % of landscapes are dominated by the agrarian landscape (agricultural areas cover more than 6 600 ha in the reference areal unit of 10 000 ha); 23.5 % of landscapes are mid-dominated (agricultural areas range from 3 300 to 6 600 ha); in 42.5 % of occurrences other landscape types are dominating (the agricultural area is smaller than 3 300 ha). In terms of diversity 37 % of the agrarian landscape hosts more than 12 different crops (or groups of crops similar from a landscape perspective); 49 % shows mid-diversity (6 to 12 crop groups); 14 % is characterised by a low diversity (less than 6 crops groups) (Figure 1 and Map 1). The index is calculated on 10 km x 10 km cells to allow upscaling results to administrative units (regional and country level). Nine landscape structure classes are shown in Table 1. 
  • Agriculture greatly influences the values of the hemeroby index. 43 % of the agrarian landscape in the EU is assigned to low to medium hemeroby levels, meaning a relatively small deviation from potential vegetation (classes 2 to 4a in Table 2), the remaining 57 % is characterised by medium-high values of of the index (classes 4b to 5b in Table 1), showing higher deviations (Figure 2 and Map 2). High Nature Value (HNV) farmland is accounted for in the 43 % share.
  • Social awareness of the agrarian landscape, measured according to indicators (surface of protected agricultural areas for ecological and/or scenic values, farm tourism and number of certified products (i.e. food and wine) linked to landscape) that describe the interest of people and society for this landscape type, is medium to high in 111 NUTS 2 regions, and medium to low in 149 NUTS 2 regions. The threshold is set at the maximum of the Gaussian distribution of values of the composite indicator in EU NUTS 2 regions, which corresponds to value 7 in Map 4. Currently the Member States which joined the EU in and after 2004 are characterised by lower scores of the indicator on social awareness than older Member States. This does not mean that their landscapes do not have high aesthetic qualities, but rather that having entered the EU in 2004 their tradition concerning i.e. EU quality schemes still has to consolidate. 

Assessment

Scientific literature accounts for a countless number of studies analysing landscapes and trying to model and characterise them. Very few examples exist, though, of landscape studies that address the whole of the European Union, and none that addresses specifically the agrarian landscape. Therefore this indicator presents a novel approach to landscape characterisation, in particular for the part concerning awareness by society. Overall, landscape is an issue for which, contrarily to many other indicators, setting thresholds is difficult or not possible at all. Therefore the contribution that change monitoring can bring is emphasised.

The methodology is harmonised throughout the Member States, but structural differences among EU regions should be taken into consideration in the analysis of results. The component on landscape structure, for example, does not provide a quality judgement on the different landscape types, but when routinely monitored will highlight areas of change, and cross-analysis with other indicators will offer the possibility for an assessment of the impact of changes on landscapes contextualised in the different EU regions. In fact the same trajectory of change (i.e. more fragmentation) could be beneficial in some cases (e.g. if patches of semi-natural vegetation fragment an intensively cultivated area) and not in others (e.g. in the urban fringe it is a sign of urban sprawl and loss of rural landscape). The same for the component on societal awareness, the ranking of regions is not linked to the notions of “good” and “bad”; it is a proxy for the awareness that societal groups have of the rural landscape according to a define set of indicators; furthermore it is not necessarily linked with aesthetic appreciation either. 

  • Regarding the landscape physical structure, as illustrated in Figure 1 such a scheme allows the identification of the main types of agrarian landscape in terms of structure. A low diversity and low dominance characterise in fact i.e. the alpine pastures in a forested context; a high dominance and low diversity represents areas with a homogeneous landscape type (wine regions, rice fields, rough grazings etc.); high dominance and high diversity is a landscape where agriculture dominates but that is also characterised by a high internal variability of crops; high diversity and low dominance is typical i.e. of the urban fringe, where agricultural fields are scattered among other land uses and include many different cultivation types. In Map 1 the classes shown in red are those where landscape is dominated by agriculture, agricultural patches (intended as areas under agricultural use) are very large (more than 6 600 ha in the reference areal unit of 10 000 ha) and composed by a wide variety of crops (i.e. intensively cultivated area in Europe mostly fall under this class). Classes shown in light orange are areas where the agrarian landscape is dominating, but these are either monocultures or areas characterised by the presence of very few crops (i.e. Irish pastureland). Classes shown in dark blue are those where agriculture is fragmented (the area of an agricultural patch is smaller that 3 300 ha in the reference areal unit of 10 000 ha) but is composed by a wide variety of crops (i.e. some urban fringes, or densely populated areas containing scattered agricultural land). Classes shown in light blue are those areas with a low dominance of agriculture (patches smaller than 3 300 ha in the reference areal unit of 10 000 ha) and a low crop variety (i.e. mountain pastures). The other five classes are intermediate situations between the four mentioned above.
  • The hemeroby index shows the picture of the impact of agricultural activities on landscapes (Map 2) and also in relation to other land uses (Map 3). The final result shows areas where intensive crop production is located, and on the other hand where the impact of human activities is lower.
  • In the final indicator, the reasons why the regions score high are very different: some have a high rate of protected agricultural area (e.g. Rhein regions and Baden-Wuttemberg), some have a high number of certified products (e.g. Dytiki Makedonia in Greece for cheese, Norte in Portugal for meat), some have a high number of farms declaring relevant revenue from tourism activities (e.g.Toscana, Tirol and Salzburg). On the other hand it can also happen that some regions (e.g. Burgenland in Austria) have a high score because they reach medium results in all indicators. The Swedish region Övre Norrland scores high because its agricultural land is contained in protected areas and the value of the indicator is normalised on the UAA, therefore small areas may get high values (meaning that society is aware of them according to the identified criteria).

Data sources and availability

Indicator definition

The landscape state and diversity indicator describes the main characteristics of the agrarian landscape, in terms of structure of the landscape, cultural influence on the potential natural vegetation due to human activities, and societal awareness of the rural landscape. 

Measurements

The indicator is structured in three components:

  • Landscape physical structure
  • Hemeroby index
  • Societal awareness of agrarian landscape

Links with other indicators

The indicator "Landscape - State and diversity" is linked with following other indicators:

Data used and methodology

The following databases were used:

The methodology applied does not differ among regions. Main constraints to obtaining consistent results are linked to data accuracy (i.e. the distribution of crop types is the result of a disaggregation procedure based on FSS statistics, LUCAS data and environmental variables; the information on number of holdings having tourism as “other gainful activity” in the FSS, and “receipts of tourism, including returns from board and lodging, campsites, cottages, riding facilities, hunting and fishing and excluding value of products produced on the holding used for catering” in FADN is not complete), and to the weighting system applied in order to include in the indicator both food certified products (number of labels) and quality wines (hectarages). Data availability also limits the descriptive capacity of the indicators (i.e. lack of information on field size and landscape elements). Some of these limitations may be solved when new information becomes available i.e. through FSS or Land Parcel Identification System (LPIS).

It is worth noting that the hemeroby index is related to land use intensity but there are slight differences between the two concepts, for example extensive grasslands have hemeroby values lower than extensive cropland because cropland includes always a mechanical action on the soil, which brings the system further away from the natural state.

The methodology used by constructing the particular components of the indicator was as follows:

Component 1 – Landscape physical structure

The indices identified to describe the physical structure are the Largest Patch Index (LPI) as a measure of agrarian landscape fragmentation in the matrix of non-agricultural background, and the number of crop categories as a measure of agrarian landscape diversity.

The Largest Patch Index (LPI) was calculated for a 10x10 km cell grid covering the EU-27. The Corine land cover 2000 (CLC2000) raster dataset was split into 10x10 km raster squares, and then reclassified into two categories: “Agriculture”, including agricultural classes and natural grasslands, and “background”, including artificial areas, natural vegetation and water. LPI was then calculated for each 10x10 km raster square using Fragstat 3.3.[1]. LPI ranges from 0 to 100 and is expressed as percentage. Following the above described protocol, LPI measures the extension of the largest agricultural patch in each cell, and thus the dominance and fragmentation of agrarian landscapes.

The number of crop categories was as well calculated for a 10x10 km cell grid covering the EU-27. Data from the CAPRI were used. The CAPRI model allocates crops and estimates their share of UAA in Homogeneous Soil Mapping Units (HSMU), consistently with statistics at NUTS 2 level. The distribution of 18 crop categories was analysed: cereals, maize, paddy rice, rapeseed, sunflower, legumes, textile fibres, other industrial crops, nurseries, flowers, vegetables, root crops, tobacco, fruits, citrus fruits, olives, grapes, grasslands. The categories have been defined according to their significance from a landscape perspective. The information at HSMU level was aggregated at the 10x10 km grid resolution, assuming that the crops available in each HSMU could be uniformly distributed within its area. Nine landscape structure classes were then created by cross combination of the two indices, as illustrated in Table 1. Such a scheme allows the identification of the main types of agrarian landscape in terms of structure.

Component 2 – Hemeroby index

The hemerobiotic state represents the magnitude of the deviation from the potential natural vegetation caused by human activities. The degree of hemeroby increases with the increase of the human influence. Gradients of human influence are assessed using a scale which normally comprises 7 degrees, in which the lowest values (ahemerob) correspond to “natural” or non-disturbed landscapes and habitats such as bogs and the highest values (metahemerob) are given to totally disturbed or “artificial” landscapes such as urban areas. In the present context the index shows the cultural influence of farming practices on landscapes and potential vegetation.

The hemeroby scale adopted in the present exercise (Paracchini and Capitani, 2011)[2] is presented in Table 2. The original scale[3] [4] has been revised in order to allocate with more detail broad categories of agricultural land uses. The latter, in fact, are associated to hemeroby levels 2 to 5, and especially grasslands and arable land have a consistent overlap in level 4. The proposed revision splits levels 4 and 5 in two parts, so that grasslands can be associated to levels 2 to 4b, ranging from “close to natural” for light management like transhumance to “relatively far from natural” when they are heavily managed and therefore composed by very few species; arable land is associated to levels 4a to 5b (“far from natural”), ranging from annual crops associated with permanent crops to cereal monocultures. In any case grasslands do not exceed level 4b, and arable land level 5b.

The index was calculated using Corine land cover 2000 and CAPRI data on N input and livestock density. Each CLC2000 class was assigned to an average degree of hemeroby on the basis of examples found in literature. Combined information on levels of inputs was derived from a reclassification of N input and livestock density data, available in the CAPRI modeling system at 1 km resolution. Thresholds were set to 30 kg/ha and 150 kg/ha for N input and to 0.5 and 1.2 LSU/ha (Livestock unit per hectare) for livestock density and three classes of low-medium-high levels of input were obtained by selecting in each 1 km cell the highest of the two levels of inputs. This information was then used to rescale CLC2000 agricultural classes to the degrees of hemeroby, assigning the medium-input class to the predefined value, and the low- and high- input levels to the two contiguous degrees. The hemeroby state for natural and semi-natural vegetation was evaluated as well in order to give the full picture not only of the impact of agricultural activities on landscapes (Map 2), but also in relation to other land uses (Map 3). The final result shows areas where intensive crop production is located, and on the other hand where the impact of human activities is lower. At this regard it should be noted that the hemeroby index is related to land use intensity but there are slight differences between the two concepts, for example grasslands have hemeroby values lower than arable areas (except in the case of sport facilities and golf courses) because cropland always includes a mechanical action on the soil, which brings the system further away from the natural state.

Component 3 – Social awareness of the agrarian landscape

The component is a proxy for the interest/perception that society has for the agrarian landscape[5]. The indicator is a composite indicator (OECD/JRC 2008), and precisely an aggregation of three indices:

The index for Quality food and wine was calculated from two different datasets: firstly, PDO and PGI products linked to landscape state and diversity were selected from the DOOR database. The selection was based on the following criteria: 1) the product itself creates a specific landscape (i.e. vineyards, olive groves, etc.); 2) the production area is characterized by a particular landscape (i.e. montados, bocages, alpine meadows, maquis, etc.); 3) the production is explicitly related to the preservation of the landscape’s characteristics; 4) the production is the result of a traditional management of rural landscape. A geo-database of the spatial distribution of selected PDO and PGI products was created at NUTS 3 level, according to the information on the production areas provided by producers. Then, the number of different certified product per NUTS 2 region was calculated. For VQPRD wines data on the cultivated surface (ha) was extracted from the Inventory of wine-growing areas, available at NUTS 2 level. The index was calculated as the surface under cultivation of quality wines produced in specified regions.

The two indices (food and wine) were standardised on the utilised agricultural area (UAA) and rescaled in the 0-10 range on the basis of the minimum and maximum values of the sample and summed.

The standardisation was necessary in order to refer the variable to the agrarian landscape, and avoiding giving low scores to regions only because they contain a small proportion of this landscape type. Even if small, when such landscape is of interest to society, this must be recognised. The equation for rescaling the values is the following:

Irescaled = (Inorm – Imin) / (Imax – Imin) * 10     (equation (1))

Where:

Irescaled is the result of the rescaling and final value of the index
Inorm is the result of the normalisation on the UAA
Imin is the minimum value of the population of Inorm calculated at NUTS 2 level
Imax is the maximum value of the population of Inorm calculated at NUTS 2 level

It is then necessary the rescale once more the results by applying this equation in order to match the range of variation of the other two components, which is 0-10.

The second index composing the societal awareness indicator is related to tourism activity in rural areas, for which data are both fragmented and incomplete at European scale. It was calculated according to two sources:

  • FSS declarations for “Tourism as other gainful activity” which refer to all activities in tourism, accommodation services, showing the holding to tourists or other groups, sport and recreation activities etc. where either land, buildings or other resources of the holding are used;
  • FADN data on “receipts of tourism, including returns from board and lodging, campsites, cottages, riding facilities, hunting and fishing and excluding value of products produced on the holding used for catering”. These data do not represent the whole touristic activity in rural areas, but they are used in this indicator because, being linked directly to farm multifunctionality, they provide a direct link between agricultural activities and tourism. Furthermore, they are available for almost all Europe, at regional resolution.

FSS data are missing for the following regions: Eastern and South Western Scotland, Highlands and Islands in the United Kingdom and Île de France in France. FADN data for Romania and Bulgaria are currently missing, and are incomplete for Spain. FSS statistic data from 2001 to 2005 are used, and for each region data are chosen from the last available date. FADN data for 2000-2008 are averaged. The two groups of data were rescaled according to the equation (1), and averaged per each NUTS 2 region (if the FADN region is larger its value is assigned to each NUTS 2 region that is comprised in the FADN region) to produce the final index.

The third index is the share of agricultural area in protected and valuable sites, specifically Natura 2000 sites, World Heritage Unesco sites related to agricultural landscape, European nationally designated areas, and category V - World Protected Areas. Many sites are included in more than one dataset, and so a unique database was built in order to avoid redundancy. Agricultural areas are extracted by CLC2000 taking into account all agricultural classes and the class “Natural grassland”. The index was calculated as the surface of agricultural area included in protected and valuable sites in each NUTS 2 region. The index in this case is standardized by CLC2000 agricultural surface, and then rescaled in the 0 -10 range by means of equation (1).

Finally, the three indices are summed up to the final indicator which ranges potentially from 0 to 20 (Map 4). The component is organised in a way that improvements are possible. Whenever data better describing or implementing one of the indices become available (e.g. surveys on rural tourism, surfaces cultivated with PDO/PGI products etc.), they can be easily added in the model. This applies as well to the dimensions in which the index is structured, if a new dimension describing the awareness of society for the agrarian landscape becomes available (e.g. press review of rural landscape in the news) or is deemed necessary, it can be plugged in the model. Of course whenever the methodology is changed, caution should be taken when comparing values of the indicator through time.

Context

Policy relevance and context

At the European level, there are or have been several policy instruments dealing with landscapes. Among these are the Convention on Biological Diversity (CDB),1992; the Pan-European Biological and Landscape Diversity Strategy (PEBLDS) until 2012, and the European Landscape Convention (ELC),2000. The Landscape Convention was adopted on 20 October 2000 by the Council of Europe’s Committee of Ministers and came into force in March 2004 (Council of Europe Treaty Series Nº176). The convention aims to encourage public authorities to adopt policies and measures at a local, regional, national and international level for protecting, managing and planning landscapes throughout Europe. One of the specific measures of the ELC requires Parties to carry out research and studies in order to identify landscapes and analyse their characteristics and the dynamics and pressures that affect them. The Rural Development policy for the programming period 2014-2020 state that resources devoted to Union Priority 4 should contribute to the “restoring, preserving and enhancing biodiversity, including in Natura 2000 areas, and in areas facing natural or other specific constraints, and high nature value farming, as well as the state of European landscapes”[6]. In particular, in the current rural development programming period Agri-environment-climate payments play an important role in encouraging farmers to apply agricultural methods compatible with the protection and improvement of the landscape and its features. But resources can also be devoted to studies and investments associated with the maintenance, restoration and upgrading of rural landscapes, or cooperation actions preserving agricultural landscapes.

Agri-environmental context

The indicator addresses agrarian landscape monitoring from different perspectives.

Landscape physical structure is linked in first instance to the mere presence of the agrarian landscape. Agriculture is the main land use in the EU, but spatial distribution of agricultural patches is uneven and the indicator is a measure of the degree to which the agrarian landscape shapes and dominates the overall landscape, which is linked to the provision of landscape as public good (when such public good is the agrarian landscape per se). The spatial pattern of marginalisation phenomena and consequent land abandonment, urban sprawling, afforestation are among the main drivers of changes in the indicator. Physical structure is also described by crop diversity, where crop groups that are relevant from a landscape perspective (which in this case includes perception besides functionality) are identified. When diversity is related to dominance, a gradient of agrarian landscape types can be described which spans from monocultures to polycultures and from absolute dominance to high fragmentation.

The hemeroby index is a measure of intensity of management, representing the pressure of agricultural practices on habitats in terms of (qualitative) distance from the natural state. It is particularly sensitive to changes from grassland (lower hemeroby value) to cropland (higher hemeroby value) or vice versa. Also in this case marginalisation/intensification phenomena are main drivers for change.

The social awareness component is a bridge between the physical landscape and the way society interacts with it. Through its three indices (tourism, quality products linked to landscape, nature/landscape protection) the indicator reacts to the societal response linked both to landscape and developments fostered by Rural Development Policy (territorial development of rural areas, enhancement of competitiveness and farm viability, sustainable management of natural resources).

It should be noted that though the indicator on landscape state and diversity is built on the basis of a self‐standing methodology, its optimal use consists in reading the information it provides in the context of the frame of which it is part, and in a monitoring routine. If periodically calculated, the indicator can highlight hotspots of changes in the rural‐agrarian landscape, and by building storylines based on the information provided by the present and other agri-environmental indicators landscape dynamics can be fully assessed.

See also

Further information

Publications

Dedicated sections

Source data for tables, figures and maps (MS Excel)

Other information

Legislation: Commission Staff working document accompanying COM(2006)508 final
Corresponding Fact sheet 32

External links

  • Publications:
  • Database:
  • Methodology
  • Other external links:

Notes

  1. ↑ McGarigal, K., Cushman, S. A., Neel, M. C., Ene, E., 2002, FRAGSTATS: Spatial pattern analysis program for categorical maps, Computer software program produced by the authors at the University of Massachusetts, Amherst., http://www.umass.edu/landeco/research/fragstats/fragstats.html, 2011/07/21.
  2. ↑ Paracchini M.L. and C. Capitani (2011). Implementation of a EU wide indicator for the rural-agrarian landscape - In support of COM(2006)508 “Development of agri-environmental indicators for monitoring the integration of environmental concerns into the Common Agricultural Policy”. EUR 25114 EN. Publications Office of the European Union
  3. ↑ Jalas, J. (1955). Hemerobe and hemechore Pflanzenarten. Ein terminologischer Reformversuch. Acta Fauna Flora Femm. 72 (11), 1–15.
  4. ↑ Steinhardt, U., Herzog, F., Lausch, A., Muller, E., Lehmann, S. (1999). Hemeroby index for landscape monitoring and evaluation. In: Pykh, Y.A., Hyatt, D.E., Lenz, R.J. (Eds.), Environmental Indices—System Analysis Approach. EOLSS Publishers, Oxford, pp. 237–254.
  5. ↑ Paracchini M.L., Capitani C., Schmidt A.M., Andersen E., Wascher D.M., Jones P.J., Simoncini R., Carvalho Ribeiro S., Griffiths G.H., Mortimer S.R., Madeira L. , Loupa Ramos I. and T. Pinto Correia (2012). Measuring societal awareness of the rural agrarian landscape: indicators and scale issues. EUR Report 25192 EN, Luxembourg: Publications Office of the European Union
  6. ↑ see Regulation (EU) No 1305/2013 of the European Parliament and of the Council of 17 December 2013 on support for rural development by the European Agricultural Fund for Rural Development (EAFRD) and repealing Council Regulation (EC) No 1698/2005.
Retrieved from "https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Archive:Agri-environmental_indicator_-_landscape_state_and_diversity&oldid=401866"