Navigation – Plan du site

AccueilRubriquesEspace, Société, Territoire2022Towards an index of biodiversity ...


Towards an index of biodiversity potential: example in an anthropised and fragmented landscape matrix

Indicateur de biodiversité potentielle : exemple dans une matrice paysagère anthropisée et fragmentée
Indicador de biodiversidad potencial: ejemplo de una matriz de paisaje antropizado y fragmentado
Guillaume Schmitt, Magalie Franchomme, Christelle Hinnewinkel et Marie Laboureur


La connaissance de la biodiversité est un enjeu majeur de la planification urbaine. L’estimation de la biodiversité potentielle des entités de la matrice paysagère mobilise souvent les outils de l’analyse spatiale et les indicateurs de l’écologie du paysage, notamment dans de vastes zones d’étude. Dans cet article, nous exposons la constitution d’un indice composite (CBPI) mobilisant les approches structurelle et fonctionnelle de l’écologie du paysage. Il est composé de 8 indicateurs (de qualité : perméabilité, fragmentation, gestion environnementale ; de forme : taille, complexité de la forme ; de configuration : contraste avec les polygones voisins, distance aux habitats les plus favorables, distance aux zones bâties). Le CBPI apparait adapté aux matrices paysagères anthropisées et fragmentées. Il peut servir dans l'identification des réseaux écologiques, dans la démarche Eviter, Réduire, Compenser et il peut être actualisé avec de nouvelles bases de données cartographiques.

Haut de page

Texte intégral


1Promoting biodiversity has become an essential element in urban planning (Fischer et al., 2018). The Convention on Biological Diversity (1992, art. 2) defines biodiversity as “the variability among living organisms from all sources including, inter alia, terrestrial, marine and other aquatic ecosystems and the ecological complexes of which they are part: this includes diversity within species, between species and of ecosystems”. Therefore, biodiversity covers three levels of genetic diversity, species diversity, ecosystem diversity and landscape diversity. At the global scale, the main factors contributing to the loss of biodiversity include the following:

  • habitat modification, fragmentation and destruction, especially through the growth of built-up areas and transport infrastructure,

  • introduction of invasive species,

  • pollution of ecosystems,

  • over-consumption of natural resources and,

  • climate change (Secretariat of the Convention on Biological Diversity, 2014).

  • 1 Modelling outputs are not always produced at the scale of political territories and their spatial r (...)

2Sustainable urban planning involves reducing the effects of these main factors, but also strengthening and creating green infrastructure, such as greenways (Walmsley, 2006). Urban planners need biodiversity indices to identify effective planning solutions for both biodiversity and development in anthropised areas (Hermy, Cornelis, 2006; Bekessy et al., 2012). According to Walz and Syrbe (2010), biodiversity indices fall into two categories: those relating to species richness based on fauna and flora inventories and those relating to ecosystem characteristics on the basis of land use. The former indicators are heterogeneous and insufficiently adapted to the needs of urban planners (in terms of perimeter and spatial resolution). Moreover, urban planners generally have little training in ecology1 (Alphandéry, Fortier, 2015). The latter indices have been more widely used, especially since the advent of landscape ecology (O’Neill et al., 1988). The landscape can be a link between scientific knowledge on fauna and flora and urban planning (Mander, Uuemaa, 2010), because the organisation of the landscape matrix influences biodiversity according to biophysical conditions and land-use patterns. Landscape metrics can describe the variety of the landscape matrix, and also have connections with species diversity and abundance (McGarigal, 2002).

3Landscape metrics are numerous and require a detailed knowledge of ecological processes (Walz, 2015). Landscape metrics for estimating the biodiversity potential of habitats have been the subject of numerous investigations and creation of synthetic indices (Kujala et al., 2015; Tarabon et al., 2020). In this article, we focus on the most widely used indices found in the literature and that can be used in a geographic information system (GIS), the preferred urban-planning tool (Vimal, Mathevet, 2011). The study area is highly anthropised and the landscapes are very fragmented. The urban planning documents for this area endeavour to strengthen connectivity by introducing a greenway (Franchomme et al., 2013), therefore drawing heavily on indices of connectivity and habitat permeability. Thus, this article focuses on biodiversity-related indices to model an anthropised and fragmented landscape matrix.

  • 2 Most of the indices have been calculated in Python language or use common spatial operators (overla (...)

4In the first part, the construction of a composite index is detailed, because it is more easily appropriated by planners than several indices. With planners in mind, we also favour the vector format within a GIS2 in the various calculations of landscape metrics (Yu et al., 2019). This method is based on spatial analysis tools and the use of knowledge of experts in ecology, law, urban planning and geography. The results and contributions of the index are presented in the second part, focusing on the possibilities of monitoring the biodiversity potential in the initial phases of urban planning projects.

5The study area comprises two intermediate French administrative divisions (known as départements, hereafter referred to in English as ‘departments’): Nord and Pas-de-Calais (12 485 km²), with a population of 4,077,866 at the last population census. Natural areas are rare in this region, having the lowest surface area among all French departments, either in absolute or relative terms (Schmitt, 2009).

Figure 1: Map of the study area and its land uses

Figure 1: Map of the study area and its land uses

Method: modeling the landscape matrix

  • 3 Six indices were calculated on the land use/cover polygons in the database presented below. Two ind (...)

6We developed a composite biodiversity potential index (CBPI) to model the landscape matrix of the study area. This index was constructed using eight indices3 of spatial quality, pattern and structure borrowed from landscape ecology (Riiters et al., 1995; Aguejdad, Hubert-Moy, 2016) and supplemented with expert consultations during the whole modeling process (Dale et al., 2019). The choice of indices considers structural connectivity (Taylor et al., 2006). It corresponds to the organisation of the landscape patches that make up the landscape mosaic. Landscape structure affects the movement of species and is a component of biodiversity (Renetzeder et al., 2010). In addition, functional connectivity is also assessed from a generic species profile (medium-sized mammals) to integrate the processes of supplementation and complementation in a fragmented landscape matrix (Ouin et al., 2004). The first process is related to the movement of a species to obtain the necessary resources in a fragmented matrix for small habitats. The second process is linked to the life cycle of the species (reproduction, migration, etc.). Figure 2 illustrates the eight indices that are component of the CBPI in the study area. It is possible to distinguish the discriminatory potential of each index. The eight indices are presented in the following sections.

Figure 2: The eight indices composing the composite biodiversity potential index (CBPI)

Figure 2: The eight indices composing the composite biodiversity potential index (CBPI)

Quality indices characterising the biodiversity of the landscape

7Three quality indices were used in this study, and involve land use, fragmentation and environmental management of each entity in the landscape matrix.

Permeability index

8The first index was developed from the latest available land-use map for Nord and Pas-de-Calais. The map is based on two sets of aerial photographs taken in 2009 at a spatial resolution of 20 cm in the visible and near-infrared regions of the electromagnetic spectrum. The map uses European Corine Land Cover (CLC) nomenclature with three nested levels and also includes a fourth level, giving a total of 54 categories to refine the distinctions within each of the following areas:

  • artificial surfaces, varying with building density and use intensity (11 categories in the CLC vs 26 in the database we used),

  • agricultural areas, varying with crop assortment and cultivation practices (4 categories vs 6),

  • natural areas, with for example a distinction between forested areas according to planting or harvest date (13 categories vs 22).

  • 4 Mustelidae (weasels, European badgers, ermines).

9With a recommended working scale of 1:25,000, a source scale of 1:15,000 and an observed minimum mapping unit of less than 100 m², the represented landscape matrix is complex and is composed of some 125,000 objects. Using a previously tested method, the various categories were grouped into six levels of increasing permeability for movements of medium-sized mammals4 common in northern France by distinguishing between areas that are difficult to cross, areas to be avoided, areas of occasional passage, substitute habitats, habitats amenable to reproduction and habitats ensuring high reproduction and survival rates (Mimet et al., 2017; Schmitt et al., 2014). This species profile was based on a generalist medium-sized mammal for several reasons. Medium-sized mammals have an ecological profile corresponding to the landscape matrix in terms of habitat size, dispersal capacity and ecosystem (Opdam et al., 2008). Their profile corresponds to a landscape species approach (Sanderson et al., 2002). These mammals are studied by managers and developers because they are sensitive to management plans and actions to promote connectivity (Calenge et al., 2015), which is the main objective of conservation planning (SRCE, Schéma Régional de Cohérence Écologique) in the study area (Weber, Allen, 2010). The regional wildlife organisation of the study area reports the presence of 41 mammal species including 22 considered ‘threatened’ according to the International Union for Conservation of Nature (IUCN). The average distance of 5 km between the biodiversity reservoirs identified in the regional land-use and conservation planning scheme (SRCE) is consistent with the movements of a medium-sized mammal (Minor, Lokkingbill, 2010).

10This typology was compared with the expert knowledge of three ecologists on three test areas where they had conducted surveys (Table 1). Ten levels of permeability to movement were selected with a wider range between levels, particularly for areas to be avoided (Kinldmann, Burel, 2008):

  • areas that are very difficult to cross (8 land-use categories, score 100),

  • areas that are difficult to cross (9 land-use categories, score 95),

  • areas to be avoided (1 land-use categorie, score 80),

  • areas of occasional passage (6 land-use categories, score 75),

  • substitute habitats (2 land-use categories, score 70),

  • better substitute habitats (2 land-use categories, score 60),

  • habitats amenable to reproduction (4 land-use categories, score 50),

  • habitats very amenable to reproduction (2 land-use categories, score 40),

  • habitats ensuring high survival rate (4 land-use categories, score 20),

  • habitats ensuring high reproduction and survival rates (12 land-use categories, score 1).

11Table 1 also contains the change in the number of polygons in the initial land cover database after its modification following the scoring of the eight indices. The number of polygons increased from 125,000 to nearly 520,000, providing a finer estimate of biodiversity potential. Table 1 also contains the mean, median and standard deviation of CBPI for all land-use categories. The highest scores correspond to the least favourable habitats and the lowest values to the most favourable habitats. Variations in CBPI for the same habitat type or land-use category are explained by spatial variations in quality, shape and configuration indices.

12Then, the indices of quality, shape and configuration were normalised from 0 to 100 to make them easier to weight (Liénard, Clergeau, 2011).

Table 1: The permeability index and the change in the composition of the landscape matrix after scoring the quality, shape and configuration indices

Table 1: The permeability index and the change in the composition of the landscape matrix after scoring the quality, shape and configuration indices

Habitat fragmentation index

13The second quality index corresponds to an assessment of the fragmentation of each occupied patch. This index was used for a practical reason: land-use maps do not show all roads (only roads exceeding 25 m in width were inventoried). Also, from a landscape ecology point of view, roads are known to have effects – in terms of road kill or noise pollution – on species diversity and richness (Forman, Alexander, 1998). The density (linear metres per km²) of all passable roads were thus calculated for each patch by weighting them by their number of lanes, their administrative classification and their width, when this information was available in large-scale reference maps (IGN). This quality index also constitutes an index of exposure to human activity (Bernier, Théau, 2013 – Figure 3). This index is easily updated, because it uses standardised, open data.

Figure 3: Illustration of the habitat fragmentation index

Figure 3: Illustration of the habitat fragmentation index

Environmental management index

14The third quality index involves the estimation of biodiversity. We originally planned to use fauna and flora inventories, but their great heterogeneity, their temporal and spatial anisotropy and their scarcity led us to choose a different approach (Alphandéry, Fortier, 2015). Thus, the environmental management index was estimated using different zoning categories used for inventory and for protected areas (Dudley, 2008). In the study area, there are 19 coexisting zoning categories or zones, which can overlap. For example, up to eight zones overlapped in the Audomarois marshes near Saint-Omer. These different zones can be defined by international law (biosphere reserves part of the Man and Biosphere programme, MAB), European law (Special Protection Areas, SPAs), national law (biological reserves), be part of regional policies (regional natural reserves) or formulated locally (‘natural areas’) identified in local land-use plans (PLU, Plan local d’urbanisme). These zones do not all have the same regulatory scope (more or less restrictive measures) and sometimes involve specific environment management methods (Natura 2000 reference documents), more or less strict zoning rules regarding constructability (natural areas and classified wooded areas in local land-use plans) or are simply areas delimited for inventories (Natural Areas of Ecological, Faunistic and Floristic Value, ZNIEFF). Therefore, the use of these zones as an environmental management index was cross-checked with experts in land-use planning, environment, geography and ecology, who sit on the scientific boards of regional natural parks or MAB reserves.

15More specifically, we first produced a map of these zones and then showed this map individually to five experts to confer with them about what these zones actually meant. These interviews showed that the number of zones for a given patch was not a sufficiently significant variable. Although an area covered by high number of zones likely indicates an area with high environmental value or stakes, zones also reflect work on land-use planning in the interest of keeping urbanisation in check or to promote tourism (Lorant-Plantier, 2014). Moreover, not all zones benefit from the same degree of conservation. For example, acquisition of public land (i.e. coastal or lakeside areas) or the investment of public funds are the manifestations of a desire to preserve the environment or of the presence of remarkable or high biodiversity (Table 2).

Table 2: The different denominated zones found in the study area and their weights based on consultation with experts for the development of a composite biodiversity potential index

Table 2: The different denominated zones found in the study area and their weights based on consultation with experts for the development of a composite biodiversity potential index

16After interviews with experts, we organised a group meeting during which each expert proposed a ranking of zones according to his/her field of knowledge. During the group meeting, IUCN's Protected Area Categories System was not satisfactory. Indeed, some areas classified with the IUCN system as high level had a lower biodiversity potential according to experts than other areas managed by the department (classified in the IUCN system with a lower level of protection). The confrontation of the different rankings helped establish a score for each zone. In the case of disagreements amidst experts, the land use was used as the final decision criterion. For example, ZNIEFF 1 (score 15) is composed equally of artificial areas, forest and semi-natural areas. ZNIEFF 2 (score 10) has one-third less forest and semi-natural areas, and more agricultural and artificial areas. Within the two inventory areas, parcels often have higher scores if they are included in national or regional reserves (Figure 4). The zoning categories of “natural regional reserves”, “national nature reserves” and “biotope decrees” were the most difficult to classify; in these areas, the consensus among experts was frequently more difficult to reach. Ultimately, the choice was made to apply the maximum score identified. This zone scoring system was used to assign each occupied patch a score from 0 to 100 according to whether it belonged to an inventory zone or to an environmental conservation zone.

Figure 4: Illustration of the environmental management index

Figure 4: Illustration of the environmental management index

Shape indices characterise fragmentation and landscape connectivity

17Indices of patch shape are frequently employed – and often amended – in landscape ecology to study landscape fragmentation and connectivity (Turner et al., 2015). Patch shape effectively influences mammal movements (Hardt, Forman, 1989). Indices are generally strongly inter-correlated and depend on patch size (Hargis et al., 1998).

Size index

18For each patch in the study area, we successively calculated the perimeter-to-area ratio, the shape index and the fractal dimension index (McGarigal et al., 2002). In particularly fragmented landscapes such as those found in northern France, the choice of a given patch shape index was based on the complementarity with the area index (Figure 5).

Figure 5: Illustration of size index in the study area: most patches have a low surface area

Figure 5: Illustration of size index in the study area: most patches have a low surface area

Shape index

19Shape indices are particularly used to describe natural habitats (forest, wetlands, etc.). The relationship between population or building density and biodiversity is well documented (Tratalos et al., 2007). The relationship between biodiversity and the shape of urban areas, or more generally habitats less favourable to biodiversity, is less frequently investigated (Flégeau et al., 2021). Urban structures have an influence on the biodiversity of nearby spaces (Tannier et al., 2012). The presence of mammals varies within urban areas and they are often more numerous in the suburban areas or where there is a significant presence of parks (Parsons et al., 2018). Considering the configuration of urban areas (private gardens on the periphery, decreasing density with distance from the centre, etc.) and the low proportion of natural areas in the study area (Figure 1), we created a shape index for the habitats defined by the permeability index. We measured the internal complexity of habitats encompassing several categories of land cover/use. The index is particularly discriminating for occasional or avoidable habitats (see Figure 10). In practice, we calculated the ratio between the number of vertices and the line length of the land-use polygons comprising these habitats (Moser et al., 2002). This index also provides a measure of edge effects between patches on species movements in relation to the quality indices and configuration indices (Figure 6).

Figure 6: Illustration of the shape index

Figure 6: Illustration of the shape index

Configuration indices reveal the pattern of habitat distribution

20We used three different configuration indices. They depict the connectivity potential of each patch with respect to its environment according to patch contiguity or proximity (Haines-Young et al., 1996).

Contrast index

  • 5 For technical reasons and to facilitate data processing and computation, all patch-related indices (...)

21The first index involves topological contiguity and the level of contrast between adjacent patches (McGarigal et al., 2002). It measures the degree of isolation, for example a forest in an urban area or, at the other extreme, a high degree of similarity with adjacent patches. An urban area with many borders on quality habitats will have a more favourable contrast index than an urban area surrounded by agricultural land. It was calculated in vector mode in GIS5 by subdividing the perimeter of each patch according to its contiguity at different degrees of permeability defined above. Subdivisions of each patch are then weighted according to the length of the shared edge and summed to give a single contrast index (Figure 7).

Figure 7: Illustration of the contrast index

Figure 7: Illustration of the contrast index

Isolation index

22The second index is related to the isolation of each patch and can be used to compare the degree of isolation of patches (Bender et al., 2003). Within a radius of 5 km corresponding to the average travelling distance for a medium-sized mammal, we calculated the number and distance of ‘habitats ensuring high reproduction and survival rates’ defined in the permeability index for each polygon (Figure 8). The average distance to the best habitats is the most discriminating variable and there is a positive correlation between proximity to biodiversity hotspots and biodiversity potential at the regional scale (Rüdisser et al., 2012).

Figure 8: Illustration of the isolation index

Figure 8: Illustration of the isolation index

Distance index

23The third index measures the distance from built-up areas defined from the land-use map. Proximity to built-up areas seems to generate avoidance behaviour in medium-sized mammals (Croci et al., 2008; Jolivet et al., 2015). In Nord and Pas-de-Calais, the impervious surfaces inventoried in the 2009 map cover 15.8%. We applied successive buffer zones to built-up areas, using a step size of 25 m, and a threshold of 100 m. With this buffer zone, the influence of built-up areas extends to 22.2% of the study area. With increasing distance from built-up areas, the distance index decreases with a threshold effect (Figure 9).

Figure 9: Illustration of influence of the distance index on the composite biodiversity potential index

Figure 9: Illustration of influence of the distance index on the composite biodiversity potential index

Weighting the indices to construct a composite biodiversity potential index

24After this initial preparatory phase, the landscape matrix was composed of 520,000 patches and the weight of each index was determined after consultations with ecologists, geographers and legal experts to produce a composite index of biodiversity potential (CBPI). We adopted a multidimensional index construction method based on the same principles of the permeability index and based on Dobbie and Dail (2013): use of common indices, standardisation, statistical progressiveness of weighting and aggregation, statistical testing and local studies. Compared with other multi-criterion analysis methods, this approach has the advantage of having being used on the permeability index with experts who knew the study area. Here, the weighting of the eight indices was done by the same experts who defined the permeability and environmental management index (Table 3). The same iterative method was used (individual weighting, then group meeting). However, this method is more time-consuming, because each index was gradually adjusted on a map during the group meeting and local examples (shown in Figures 3, 4, 7 and 9) were based on the experts’ knowledge of the area. In practice, the regional level of the landscape (permeability, environmental management) is favoured in the prioritisation of indices (Noss, 1990).

Table 3: Weighting of the quality, shape and configuration (exposure) indices to construct a composite biodiversity potential index (CBPI), scaled from 0 to 100.

Table 3: Weighting of the quality, shape and configuration (exposure) indices to construct a composite biodiversity potential index (CBPI), scaled from 0 to 100.

25The variations of the indices composing the CBPI according to the degrees of permeability thus illustrate the heterogeneity of the landscape matrix (Figure 10). The fragmentation index is highest for the least favourable habitats. It distinguishes the most suitable habitats according to the density of the road network. The management index refined the CBPI for the most suitable habitats. In addition, it gave a positive value to artificialised areas with a biodiversity protection system (quarry pit with the presence of endangered animals, etc.). The size index (5% of the CBPI) scores the largest habitats in relation to the smallest, in particular for the “better substitute habitats”. The shape index complements the size index. It is highest for habitats containing urban spaces and quite clearly separates the most suitable habitats according to their shape. The contrast index reduces the CBPI of the most suitable habitats with many boundaries with the least favourable habitats. Conversely, the potential of occasional habitats is enhanced when there are adjacent forests, for example. The isolation index distinguishes among the most favourable habitats according to their respective spacing. In contrast, and especially for the least favourable small habitats, the CBPI is enhanced according to the number and distance to the habitats of better quality. The distance index provides nuances in the description of the potential of habitat edges. The CBPI has an increasing average score from ‘habitats ensuring high reproduction and survival rates’ to ‘areas that are very difficult to cross’, with the exception of ‘areas to be avoided’ which correspond solely to ‘permanent crops’. The high average size of ‘permanent crops’ and the evolution of the number of polygons (+ 800% - Table 1) explain the score and deviations of the CBPI. More generally, the differences in CBPI illustrate its ability to estimate the biodiversity potential of a patch according to its qualities, shape and configuration.

Figure 10: Index statistics according to the level of permeability

Figure 10: Index statistics according to the level of permeability

Results and discussion

26The CBPI describes the low biodiversity potential of the Nord and Pas de Calais departments (Figure 11). The low values correspond to the urban framework (port area of Dunkerque, Lille metropolitan area, conurbation between Béthune and Valenciennes). The intermediate values correspond to agricultural areas of permanent crops, as in the example of the south of Arras, with a significant difference between populiculture and grassland areas in the south-east of the department of Nord. Areas of high potential correspond to non-industrialised coastlines (estuary, etc.), large forests (Desvres Forest, Saint-Amand and Mormal), as well as relict wetlands (between Douai and Cambrai in the Sensée Valley). The high density of the transport infrastructure network is particularly visible (motorways and railways between Lille and Paris) and separates areas of medium and low potential (coastal motorway south of Boulogne-sur-Mer). At a finer scale, there are many juxtapositions of high and low potential patches, particularly between urban agglomerations (between Lille and Lens), but also in the natural areas mentioned above.

  • 6 Some land-use categories involved forest age and monospecific stands.

27The CBPI is part of a line of biodiversity potential indices and in particular that developed by Larrieu and Gonin (2012) and used in Clevenot et al. (2017) on a forested area in the Greater Paris region. By definition, this index is multifactorial in its assessment of habitat biodiversity in terms of abundance, richness and functionality. It also integrates relatively different management practices and uses, and does not depend on any initial reference state. It is different from previous indices insofar as it neglects vegetation stratification and age6, is not based on field surveys and it can be extended to types of habitat other than forest areas. It describes the biodiversity potential of environments that are particularly fragmented and overlap with dense urban areas such as parks and gardens (Pellissier et al., 2014). It can also illustrate different types of habitat by focusing in particular on the potential connectivity and biodiversity of each element of the landscape matrix. Each patch of the landscape matrix is a combined function of:

  • its shape and size, thus avoiding the spread of homogeneous values over the entire study area (for example, not all grasslands have the same connectivity or biodiversity potential),

  • factors related to structural connectivity (a forest enclosed by artificial areas has a low biodiversity potential unless it is likely to be an effective biodiversity reservoir),

  • environmental management factors (a classified and closed wetland has a high biodiversity potential) and

  • more functional factors by using a landscape species profile in the determination of permeability scores (index with the greatest impact on the CBPI).

Figure 11: Illustration of the composite biodiversity potential index (CBPI)

Figure 11: Illustration of the composite biodiversity potential index (CBPI)

28The contributions and limitations of CBPI involve either the biodiversity aspect of the indices or are of a technical nature.

29By using indices that convey structural information and functional connectivity with respect to a generalist mammal species, the CBPI is a generic index of biodiversity potential in particularly fragmented and anthropised spaces. It provides a general overview of biodiversity potentials as a complement to naturalist inventories, which are generally scattered and specific to a few species. It assigns biodiversity potential to a habitat type that varies according to the configuration, environmental management, quality and shape of each ecological patch. The CBPI thus combines the ‘habitat’ and ‘species’ approaches classically used in ecological connectivity studies (Locquet, Clauzel, 2018) with a ‘locality’ approach that integrates proximity relationships within the landscape matrix. It can be used at the beginning of the initial diagnosis of the environment in urban planning schemes. It can also be used as a map-based mediation tool between the different local stakeholders: ecologists, mayors, planners and citizens (MacEachren, 2004).

30On a technical level, CBPI was carried out in vector mode within a GIS. Local authority agencies use this tool and this format more frequently used than the raster format (classically used in landscape ecology) (Vimal, Mathevet, 2011). Production methods and techniques are thus more easily adopted. Nevertheless, and despite substantial literature, the various calculations can be long to implement, especially when using complex databases. It is often a question of finding a compromise between the degree of precision, the calculation time and the computational capacities (Foltête et al., 2012). The spatial resolution of the landscape matrix in this study, however, allows analyses to be conducted from regional to local scales in a unified reference frame. This unification can help to limit misinterpretation and increase sharing between different administrative levels (Alphandéry, Fortier, 2015). CBPI also uses expert knowledge in a diagnostic approach shared between several disciplinary fields. This type of approach can be transposed during the development of urban planning tools. In terms of mediation between experts, the scoring method (e.g. value of 2 for a Natura 2000 site and value of 1 for a ZNIEFF) proved to be less efficient than the one used for CBPI with a finer scoring method (from 0 to 100).

  • 7 The value of the Local Moran's I index is calculated from 499 permutations. A 95% threshold is used (...)

31Finally, CBPI mapping can be a tool for ecological monitoring from a land-use planning perspective. For instance, it can be applied in the GIS of institutions that oversee compliance with building regulations. The CBPI can contribute to enhancing ecological avoidance, reduction or compensation approaches generally based on habitat type along with a few proximity criteria (Müller et al., 2007). For example, geostatistical tools make it possible to study proximity relationships. In his review of the literature on hotspots, Schröter (2015) identifies four commonly used methods: top richest cells, richness, intensity and spatial clustering. The latter is frequently employed in fragmented landscape matrices. The Gi * statistic identies high concentrations of polygons with high or low values within a specied distance (Getis, Ord, 1992). We followed a stepwise approach to determine the average distance to maximise the spatial correlation. The 5 km distance corresponds to the distance travelled by a medium-sized mammal and gives the highest z-score (Minor, Lookingbill, 2010). The statistical model was adjusted by increasing the control points7 and considering multiple tests and spatial dependence (Caldas de Castro, Singer, 2006). Figure 12 first distinguishes between areas with high (pale blue) and low (light red) biodiversity potential (Schröter, 2015). It also separates areas with high potential including entities with low potential (deep red) from areas with low potential including entities with high potential (deep blue). The impacts of urban planning projects on the landscape matrix can thus be further anticipated.

Figure 12: Biodiversity potential hot and cold spots

Figure 12: Biodiversity potential hot and cold spots

32In areas of low potential (51.4% of the study area), ecological compensation actions may be preferred to strengthen ecological continuities or restore environments. In the ‘low-potential entities in a high-potential area’ category (11.1%), priority may be given to measures that mitigate and reduce negative effects on biodiversity. In low-potential areas containing high-potential entities (11.3%), spatial planning projects should also strive to preserve relictual natural areas, as in areas with high potential (16.4%) by avoiding damage to biodiversity.


33The CBPI is a multifactorial index of biodiversity potential. It is composed of quality, shape and configuration indices from landscape ecology. It combines a structural and functional approach, focusing, in this article, on a generalist medium-sized mammal. It can be recalculated to (i) measure the effects of urban planning projects on biodiversity potential and (ii) provide a tool for monitoring biodiversity over time. It can also be a reference in the determination of roughness coefficients (cost of passage) in the identification of ecological corridors by spatial analysis (distance cost according to species and patches of the landscape matrix). Finally, it can be readily updated following the creation of new databases describing the coverage, use, morphology and characteristics of the landscape on a smaller scale8.

Haut de page


Aguejdad R., Hubert-Moy L., "Suivi de l’artificialisation du territoire en milieu urbain par télédétection et à l’aide de métriques paysagères. Application à une agglomération de taille moyenne, Rennes Métropole", Cybergeo : European Journal of Geography, No.766.

Alphandéry P., Fortier A., 2015, "Les données naturalistes à l'épreuve de la transparence", Etudes rurales, Vol. 1, No.195, 127-144.

Bekessy S. A., White M., Gordon A., Moilanen A., Mccarthy M. A., Wintle B. A., 2012, "Transparent planning for biodiversity and development in the urban fringe", Landscape and Urban Planning Volume 108, No.2–4, 140-149.

Bender, D. J., Tischendorf, L., Fahrig, L., 2003, "Using patch isolation metrics to predict animal movement in binary landscapes", Landscape Ecology, Vol.18, No.1, 17-39.

Bernier A., Théau J., "Modélisation de réseaux écologiques et impacts des choix méthodologiques sur leur configuration spatiale : analyse de cas en Estrie (Québec, Canada)", VertigO - la revue électronique en sciences de l'environnement [En ligne], Vol. 13 No.2, 06 octobre 2013. URL :

Bigard C., Thiriet P., Piochd S. et Thompson J. D., 2020. "Strategic landscape-scale planning to improve mitigation hierarchy implementation: An empirical case study in Mediterranean France", Land Use Policy, Volume 90.

Caldas de Castro, M. Singer B., 2006, "Controlling the False Discovery Rate: A New Application to Account for Multiple and Dependent Test in Local Statistics of Spatial Association", Geographical Analysis, No.38, 180-208.

Calenge C., Chadoeuf J., Giraud C., Huet S., Julliard R., Monestiez P., Piffady J., Pinaud D. & Ruette S., 2015, "The spatial distribution of Mustelidae in France", PLoS ONE Vol.10, No.3.

Clauzel C., Bonnevalle C., 2019, "Apports de la modélisation spatiale pour la gestion de la trame verte et bleue", Cybergeo : European Journal of Geography, No.900.

Clevenot L., De Chastenet C., Frascaria N., Jacob P., Raymond R., Simon L., Pech P., 2017, "Do Linear Transport Infrastructures provide a potential corridor for urban biodiversity? Case study in Greater Paris, France", Cybergeo : European Journal of Geography, No.803.

Croci S., Butet A., Georges A., Aguejdad R., Clergeau P., 2008, "Small urban woodlands as biodiversity conservation hot-spot: a multi-taxon approach", Landscape Ecology, Vol.23, 1171-1186.

Dale V., Kline K., Parish E., EichlerS, 2019, "Engaging stakeholders to assess landscape sustainability", Landscape Ecology, Vol.34, No.6, 1199–1218.

Dobbie M.J., Dail D., 2013, "Robustness and sensitivity of weighting and aggregation in constructing composite indices", Ecological Indicators, Vol.29, 270-277.

Dudley N., 2008, Lignes directrices pour l'application des catégories de gestion aux aires protégées. IUCN.

Fischer L. K., Honold J., Cvejić R., Delshammar T., Hilbert S., Lafortezza R., Nastran M., Busse Nielsen A., Pintar M., van der Jagt A. P.N., Kowarik I., 2018, "Beyond green: Broad support for biodiversity in multicultural European cities ", Global Environmental Change, Vol.49, 35-45.

Flégeau M., Soubelet H., Carré S., Barot S., Lagurgue X., Bognon S., Clergeau P., 2021, "What evidence exists on the possible effects of urban forms on terrestrial biodiversity in western cities? A systematic map protocol", Environmental Evidence, Vol.10, No.1.

Foltête J.-C., Clauzel C., Vuidel G., 2012, "A software tool dedicated to the modelling of landscape networks", Environmental Modelling and Software, Vol.38, 316 -327.

Forman R.T.T., Alexander L.E., 1998, "Roads and their major ecological effects", Annual Review of Ecology and Systematics, Vol.29, 207-231.

Franchomme M, Bonnin M. et Hinnewinkel C., 2013, "La biodiversité « aménage-t-elle » les territoires ? Vers une écologisation des territoires", Développement durable et territoires,

Getis A., Ord J.K., 1992, "The analysis of spatial association by use of distance statistics", Geograph Anal, Vol.24, No.3, 189–206.

Gustafson E.J., Parker G.R., 1992, "Relationship between landcover proportion and indices of landscape spatial pattern", Landscape Ecology, Vol.7, 101-110.

Haines-Young R., Chopping M., 1996, "Quantifying landscape structure: a review of landscape indices and their application to forested landscapes", Progress in physical geography: Earth and Environment, Vol.20, No.4, 418-445.

Hardt R. A., Forman R. T., 1989, "Boundary form effects on woody colonization of reclaimed surface mines", Ecology, Vol.70, No.5, 1252-1260.

Hargis C.D., Bissonette J.A., David J.L., 1998, "The behavior of landscape metrics commonly used in the study of habitat fragmentation", Landscape Ecology, Vol.13, No.3, 167-186.

Hermy M., Cornelis J., 2006, "Towards a monitoring method and a number of multifaceted and hierarchical biodiversity indicators for urban and suburban parks", Landscape and Urban Planning, Vol.49, No.3–4, 149-162.

Jolivet L., Cohen M., Ruas A., "Évaluation des conséquences d’aménagements d’infrastructures sur les déplacements d’animaux. Définition et expérimentation d’un modèle de simulation agent", Cybergeo : European Journal of Geography, No.712.

Kindlmann P., Burel F., 2008, "Connectivity measures: a review", Landscape ecology, Vol.23, 879-890.

Kujala H., Whitehead A.L., Morris W.K. et Wintle B.A., 2015. "Towards strategic offsetting of biodiversity loss using spatial prioritization concepts and tools: A case study on mining impacts in Australia", Biological Conservation, No.192, 513-521.

Larrieu L., Gonin P., 2008, "L’Indice de Biodiversité Potentielle (IBP) : une méthode simple et rapide pour évaluer la biodiversité potentielle des peuplements forestiers", Revue Forestière Française, Vol.6, 727-748.

Liénard S., Clergeau P., "Trame Verte et Bleue : Utilisation des cartes d’occupation du sol pour une première approche qualitative de la biodiversité", Cybergeo : European Journal of Geography, No.519.

Locquet A. and Clauzel C., "Identification et caractérisation de la trame verte et bleue du PNR des Ardennes : comparaison des approches par habitat et par perméabilité des milieu", Cybergeo : European Journal of Geography, No. 877.

Lorant-Plantier E., "Un indicateur global d’évaluation de la vulnérabilité écologique du milieu le long de grandes infrastructures linéaires : Pourquoi ? Comment ?", Cybergeo : European Journal of Geography, No.680.

MacEachren A.M., 2004, How Maps Work. Representation, Visualization, and Design, New York, Guilford Press.

Mander U., Uuemaa E.,2010, " andscape assessment for sustainable planning", Ecological Indicators, Vol.10, No.1, 1-3.

McGarigal K., 2002, "Landscape pattern metrics". In A.H. El-Shaarawi, W.W. Piegorsch (Eds.), Encyclopedia of Environmetrics, Wiley, Chichester (2002), 1135-1142.

McGarigal K., Cushman S.A., Neel M.C., Ene E., 2002, FRAGSTATS: spatial pattern analysis program for categorical maps. Computer software program, University of Massachusetts, Amherst.

Mimet A., Clauzel C., Foltête J.C., 2016, "Locating wildlife crossings for multispecies connectivity across linear infrastructures", Landscape Ecology, Vol.31, No.9, 1955-1973

Minor E.S., Lookingbill T.R., 2010, "A Multiscale Network Analysis of Protected-Area Connectivity for Mammals in the United States", Conservation Biology, Vol.24, No.6, 1549–1558.

Moser M., Zechmeister H., Plutzar C., Sauberer N., Wrbka T. et Grabherr G., 2002, "Landscape patch shape complexity as an effective measure for plant species richness in rural landscapes ", Landscape Ecology, No.17, 657–669.

Müller A., Legendre T., Lemaître V., Darses O., 2017, La séquence « éviter, réduire et compenser », un dispositif consolidé, Paris, ministère de l’Environnement, de l’Énergie et de la Mer.

Noss, R.F, 1990, "Indicators for Monitoring Biodiversity: A Hierarchical Approach", Conservation Biology, Vol.4, No.4, 355-364.

O’Neill R.V., Milne B.T., Turner M.G., Gardner R.H., 1988, "Resource utilization scales and landscape pattern", Landscape Ecology, Vol.2, No.1, 63-69.

Opdam P., Pouwels R., van Rooij S., Steingröver E., Vos C.C., 2008, "Setting biodiversity targets in participatory regional planning: Introducing ecoproiles", Ecology and Society, Vol.13, No.1, 20.

Ouin A., Aviron S., Dover J., & Burel F. 2004. Complementation / supplementation of resources for butteries in agricultural landscapes, Agriculture Ecosystems & Environment, No.103, 473-479.

Parsons A. W., Forrester T., Baker-Whatton M. C., McShea W. J., Rota C. T., Schuttler S. G., Millspaugh J. J. & Kays R., 2018, "Mammal communities are larger and more diverse in moderately developed areas", eLife, No.7.

Pellissier V., Cohen M., Boulay A., Clergeau P., 2012, "Birds are also sensitive to landscape composition and configuration within the city centre", Landscape and Urban Planning, Vol.104, 181-188.

Renetzeder C., Schindler S., Peterseil J., Prinza M. A., Mücher S., Wrbka T., 2010, "Can we measure ecological sustainability? Landscape pattern as an indicator for naturalness and land use intensity at regional, national and European level", Ecological Indicators, Vol.10, No.1, 39-48.

Riiters K.H., O’Neill R.V., Hunsaker C.T., Wickham J.D., Yankee D.H., Timmins S.P., Jones K.B., Jackson B.L., 1995, "A factor analysis of landscape pattern and structure metrics", Landscape Ecology, Vol.10, No.1, 23-39.

Rüdisser J., Tasser E. et Tappeiner U., 2012, "Distance to nature—A new biodiversity relevant environmental indicator set at the landscape level", Ecological Indicators, Vol.15, No.1, 208–216.

Sanderson E.W., Redford K.H., Vedder A., Coppolillo P.B., Ward SE., 2002, "A conceptuals model for concervation planning based on landscape species requirements". Landscape and Urban Planning, Vol.58, 41-56.

Schmitt G., 2009, Valeurs et usages de l'espace : approches méthodologiques des dynamiques foncières dans la région Nord-Pas de Calais, Thèse de doctorat en géographie, aménagement de l’espace et urbanisme, Villeneuve-d’Ascq, Université de Lille 1.

Schmitt G., Franchomme M., Hinnewinkel C., 2014, Réseau électrique et réseau écologique : contraintes et opportunités dans un paysage fragmenté (Nord-Pas de Calais), Projets de paysage, le 24/09/2014. URL :

Schröter M., 2015, "Spatial prioritisation for conserving ecosystem services: comparing hotspots with heuristic optimisation". Landscape Ecology, Vol.31, No. 2, 431–450.

Secretariat of the Convention on Biological Diversity, 2014, Global Biodiversity Outlook. A mid-term assessment of progress towards the implementation of the Strategic Plan for Biodiversity 2011-2020. Montréal, 155 p.

Tannier C., Foltête J.-C., Girardet X., 2012, "Assessing the capacity of different urban forms to preserve the connectivity of ecological habitats", Landscape and Urban Planning, Elsevier, Vol.105, No.1-2, 128-139.

Tarabona S., Bergès L. Dutoitb T., Isselin-Nondedeu F., 2020. "Integrating a landscape connectivity approach into mitigation hierarchy planning by anticipating urban dynamics", Landscape and Urban Planning, Vol.202.

Taylor P.D., Fahrig L. & With K.A., 2006. "Landscape connectivity: a return to the basics", in Crooks K.R. & Sanjayan M. (Eds.), Connectivity Conservation, Cambridge University Press, Cambridge, 29-43.

Tratalos J. Fuller, R., Warren P. et Davies R., 2007. "Urban form, biodiversity potential and ecosystem services", Landscape and Urban Planning, No.83, 308-317.

Turner M.G., Gardner R.H., 2015, Landscape Ecology in Theory and Practice. Pattern and Process, New York, Springer-Verlag.

Vimal R., Mathevet R., "La carte et le territoire : le réseau écologique à l’épreuve de l’assemblée cartographique", Cybergeo : European Journal of Geography, No.572.

Walmsley A., 2006, "Greenways: multiplying and diversifying in the 21st century". Landscape and Urban Planning, Vol.76, No.1–4, 252-290.

Walz U., 2015, "Indicators to monitor the structural diversity of landscapes". Ecological Modelling, Vol.295, 88-106.

Walz U., Syrbe R. U., 2010, "Linking landscape structure and biodiversity". Ecological Indicators, Vol.31, 1-5.

Weber T.C., Allen W.L., 2010, "Beyond on-site mitigation: An integrated, multi-scale approach to environmental mitigation and stewardship for transportation projects", Landscape and Urban Planning, Vol.96, No.4, 240-256.

Yu M., Huang Y., Cheng X., Tian J., 2019, "An ArcMap plug-in for calculating landscape metrics of vector data", Ecological Informatics, Vol.50, 207-219.

Haut de page


1 Modelling outputs are not always produced at the scale of political territories and their spatial resolution is not always compatible with spatial planning tools. Generally, they are applied in a 'project-by-project approach' (Bigard et al., 2020 - p. 8).

2 Most of the indices have been calculated in Python language or use common spatial operators (overlap, join, etc.)

3 Six indices were calculated on the land use/cover polygons in the database presented below. Two indices (shape and isolation) were calculated from the grouping of these polygons according to their adjacency and permeability index value or score.

4 Mustelidae (weasels, European badgers, ermines).

5 For technical reasons and to facilitate data processing and computation, all patch-related indices were calculated in vector mode.

6 Some land-use categories involved forest age and monospecific stands.

7 The value of the Local Moran's I index is calculated from 499 permutations. A 95% threshold is used to identify hot and cold spots.


Haut de page

Table des illustrations

Titre Figure 1: Map of the study area and its land uses
Fichier image/png, 697k
Titre Figure 2: The eight indices composing the composite biodiversity potential index (CBPI)
Fichier image/jpeg, 2,3M
Titre Table 1: The permeability index and the change in the composition of the landscape matrix after scoring the quality, shape and configuration indices
Fichier image/png, 263k
Titre Figure 3: Illustration of the habitat fragmentation index
Fichier image/jpeg, 496k
Titre Table 2: The different denominated zones found in the study area and their weights based on consultation with experts for the development of a composite biodiversity potential index
Fichier image/png, 158k
Titre Figure 4: Illustration of the environmental management index
Fichier image/jpeg, 436k
Titre Figure 5: Illustration of size index in the study area: most patches have a low surface area
Fichier image/jpeg, 168k
Titre Figure 6: Illustration of the shape index
Fichier image/jpeg, 224k
Titre Figure 7: Illustration of the contrast index
Fichier image/jpeg, 1,2M
Titre Figure 8: Illustration of the isolation index
Fichier image/jpeg, 244k
Titre Figure 9: Illustration of influence of the distance index on the composite biodiversity potential index
Fichier image/jpeg, 420k
Titre Table 3: Weighting of the quality, shape and configuration (exposure) indices to construct a composite biodiversity potential index (CBPI), scaled from 0 to 100.
Fichier image/jpeg, 176k
Titre Figure 10: Index statistics according to the level of permeability
Fichier image/png, 76k
Titre Figure 11: Illustration of the composite biodiversity potential index (CBPI)
Fichier image/jpeg, 1,1M
Titre Figure 12: Biodiversity potential hot and cold spots
Fichier image/png, 356k
Haut de page

Pour citer cet article

Référence électronique

Guillaume Schmitt, Magalie Franchomme, Christelle Hinnewinkel et Marie Laboureur, « Towards an index of biodiversity potential: example in an anthropised and fragmented landscape matrix », Cybergeo: European Journal of Geography [En ligne], Espace, Société, Territoire, document 1021, mis en ligne le 01 juillet 2022, consulté le 16 août 2022. URL : ; DOI :

Haut de page


Guillaume Schmitt

Maître de conférences
Laboratoire de Recherche Sociétés & Humanités (LARSH), Université Polytechnique Hauts-de-France (UPHF), France

Magalie Franchomme

Maître de conférences
Univ. Lille, Univ. Littoral Côte d’Opale, ULR 4477 - TVES - Territoires Villes Environnement & Société, F-59000 Lille, France

Christelle Hinnewinkel

Maître de conférences
Univ. Lille, Univ. Littoral Côte d’Opale, ULR 4477 - TVES - Territoires Villes Environnement & Société, F-59000 Lille, France

Marie Laboureur

Ingénieure d’étude
Univ. Lille, Univ. Littoral Côte d’Opale, ULR 4477 - TVES - Territoires Villes Environnement & Société, F-59000 Lille, France

Haut de page

Droits d’auteur

Licence Creative Commons
La revue Cybergeo est mise à disposition selon les termes de la Licence Creative Commons Attribution 4.0 International.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search