Biodiversity in environmental assessment: tools for impact prediction

(1)

B IODIVERSITY IN ENVIRONMENTAL ASSESSMENT

- TOOLS FOR IMPACT PREDICTION

Mikael Gontier

March 2005

TRITA-LWR.LIC 2025 ISSN 1650-8629

ISRN KTH/LWR/LIC 2025-SE ISBN 91-7178-013-0

(2)

Mikael Gontier TRITA LWR.LIC 2025

ii

(3)

PREFACE

The research project “Prediction tools for biodiversity in environmental impact assessment”

(PROBE) is part of the Conservation Chain (Naturvårdskedjan) research program. This program is coordinated by the Swedish Biodiversity Centre (CBM) in Uppsala and funded by the Swedish Environmental Protection Agency (Naturvårdsverket). The research program also includes a research school organising courses and seminars answering the needs of PhD students participat- ing in the program. I wish hereby to thank the coordinators of the Conservation Chain for their flawless organisation.

My deepest gratitude goes to my supervisors, Berit Balfors and Ulla Mörtberg, for enrolling me in the project and for their commitment and involvement. I am also thankful to my reference group comprised of Sonia Eriksson, Bette Malmros, Anders Sjölund, Lars-Göran Mattsson, Per Sjögren-Gulve and Torbjörn Ebenhard for the fruitful discussions that we have shared. Thanks to my colleagues at the Department of Land and Water Resources Engineering for the friendly lunch and “fika” discussions. A special thank you to Bo Olofsson who introduced me to the research world. Thanks also to Duncan McConnachie for the correction of the English. All re- maining spelling mistakes are the responsibility of the author.

I am also grateful to my and Judit’s family, even though not directly involved in my research, they have always been a great and unconditional support. Finally, Judit, there is so much to say about you that I do not know how to express the way you changed my life…

Stockholm, March 2005 Mikael Gontier

iii iii

(4)

iv

(5)

ABSTRACT

Urbanisation and infrastructure developments impact on the surrounding natural environment and threaten biodiversity. The fragmentation of natural habitats in particular is a major obstacle for the preservation of biodiversity in a long-term perspective. In the planning process, both the environmental impact assessment and strategic environmental assessment processes play a central role in the identification and prediction of impacts on biodiversity. At the same time, the development of GIS technologies and GIS-based ecological models offer new perspectives in the elaboration of predictions. In order to analyse current practices and identify the need for improvements in the environmental impact process, a review of environmental impact assessment reports was carried out. Further, a review of existing GIS methods and GIS-based ecological models is presented. The results of the review of environmental impact assessment reports show a lack of predictions in current biodiversity assessments. These asssessments often concentrate on impacts at the local scale, failing to consider large-scale and widespread impacts at the ecosystem and landscape levels. The review of GIS methods and GIS-based ecological models demon- strate the possibility to generate quantitative predictions for a specific area as well as for it’s surrounding environment. At the same time, the flexibility and reproducibility of such methods would allow predictions to be made for different alternatives or scenarios, therefore providing decision makers with relevant information of potential impacts on biodiversity. This would, in turn, result in an improved integration of biodiversity issues in physical planning and contribute to a sustainable development.

Key words: Environmental impact assessment; Strategic environmental assessment; Biodiversity assessment; Prediction tools; GIS; GIS-base ecological models

v v

(6)

vi

(7)

LIST OF PAPERS

I. Gontier, M., Balfors, B., Mörtberg, U.M. Biodiversity in environmental assessment – current practice and tools for prediction. Submitted March 2005 to Environmental Impact Assessment Review.

II. Gontier, M., 2005. Integrating landscape ecology in environmental impact assessment using GIS and ecological modelling. In: Tress, B., Tress, G., Fry, G., Opdam, P. (eds.) 2005. From landscape research to landscape planning: Aspects of integration, educa- tion and application. Springer. In press.

III. Balfors, B., Mörtberg, U.M., Brokking, P., Gontier, M. Impacts of region-wide urban development on biodiversity in strategic environmental assessment. Under revision, submitted November 2004 to Journal of Environmental Assessment Policy and Man- agement.

vii vii

(8)

viii

(9)

TABLE OF CONTENT

PREFACE... III ABSTRACT... V LIST OF PAPERS ...VII TABLE OF CONTENT ... IX

INTRODUCTION...2

Objectives of the thesis... 2

Organisation of the thesis ... 2

METHODS...2

Literature review... 2

EIA review... 3

PROTECTION OF BIODIVERSITY ...3

THREATS TO BIODIVERSITY ...4

EIA AND SEA: LEGISLATIVE TOOLS FOR THE PROTECTION OF BIODIVERSITY ...5

EIA review... 5

PREDICTION TOOLS FOR BIODIVERSITY ASSESSMENT ...7

GIS in biodiversity assessment... 8

GIS-based ecological models ... 8

DISCUSSION AND CONCLUSION ... 10

FUTURE RESEARCH ... 13

REFERENCE LIST... 14

OTHER REFERENCES... 16

ix ix

(10)

1

(11)

INTRODUCTION

Urbanisation and the development of infrastructure increase pressure on the environment. Habitat loss and fragmentation resulting from urbanisation or infrastructure developments are major threats to biodiversity (Trocmé et al., 2002; Ricketts and Im- hoff, 2003). The consciousness on biodiversity has arisen during the last decades, which is illustrated by the ratification of conventions (e.g. Convention on Biodiversity, CBD, UNCED, 1992; RAMSAR, UNESCO, 1971) at the international level and the adoption of new regulations or policies at the national and regional levels (e.g. the Swedish Envi- ronmental Quality Objectives, Gov. Bill, 1997/98:145).

Some of the tools available to predict and assess impacts from anthropogenic development are environmental impact assessment (EIA) and upcoming strategic environmental assessment (SEA). The consideration of impacts on biodiversity is part of the scope of both EIA and SEA. Several authors have discussed the quality of the biodiversity assessment within EIA (e.g. Treweek et al., 1993; Thompson et al., 1997; Byron et al., 2000; Atkinson et al., 2000). In particular, it has been shown that there is a lack of adequate methodologies for systematic and quantifiable predictions of impacts on biodiversity (Treweek et al., 1993; Thompson et al., 1997; Byron et al., 2000; Atkinson et al., 2000, Geneletti, 2002), which in turn does not allow fair and rational comparisons to be made between different alternatives of a project. Further, GIS technologies and GIS- based ecological modelling have been developing, which propose methods to assess the spatial distribution of biodiversity components and to produce habitat suitability maps.

However, the relevance and potential implementation of GIS-based ecological models for EIA and SEA is yet to be demonstrated.

Objectives of the thesis

The overall aim of the project was to study the need and possibilities for implementing prediction tools in order to assess the impacts on biodiversity from changes in the

landscape due to urbanisation and infrastructure developments.

The objectives of the study were:

• to review the state of the art in biodiversity assessment within the EIA process and to identify some of the current shortcomings in the field.

• to review GIS tools in general and GIS- based ecological models in particular with respect to their potential implementation in the EIA and SEA processes.

• to study potential improvements in the prediction of impacts on biodiversity caused by urban and infrastructure developments through the use of GIS-based methods and ecological models..

Organisation of the thesis

The first section of this thesis provides de- tails on the methods used to carry out the study, followed by results of the literature review. The literature review begins with a presentation of the international and national contexts for protection of biodiversity and continues with information on the threats to biodiversity, particularly in the case of urbanisation and infrastructure developments.

The section following introduces the role of EIA and SEA for protection of biodiversity and continues with the results of the review of EIA reports, also called environmental impact statements (EISs). Further, results on the review of GIS methods and GIS-based ecological models are presented. Finally the thesis ends with a discussion of the results and some concluding remarks on the potential improvements within biodiversity assessments.

METHODS

Literature review

A literature review was conducted on:

- Ecological/biodiversity assessment - GIS and EIA/SEA

- GIS-based ecological modelling - (Integrated assessment)

2

(12)

Emphasis in the literature review has been placed on the experiences of ecological modelling implemented in a GIS environment and the use of GIS in biodiversity assessment in general.

EIA review

A review of 38 EISs was performed (Paper I). The EISs were selected from Sweden, France, England and Ireland. All four countries share a common EIA legislation, which is derived from the EU directive on EIA (OJ, 1985). While these countries have different histories when it comes to the implementation of the EIA process, implying dissimilari- ties in the practice of EIA, each, however has to comply with the requirements of the directive.

The EISs reviewed were selected in order to obtain a high degree of homogeneity in the database, allowing for an effective comparison to be made. Only road and railways projects were chosen, with the majority being road projects (36). All the EISs were pub- lished post 1999, after the EIA directive was implemented by Sweden, whereas the other three countries had implemented the directive earlier. A final selection criterion was that the projects’ length should be longer than five kilometres, which was adhered to with the exception of three projects measuring 1.5, 3.1 and 4.1 kilometres respectively (Figure 1).

Two significant limitations were experienced in the data collection process. Firstly, access to suitable EIS was problematic (limiting the number of reports collected), and secondly, language issues meant that reports could only

be selected from French, English and Swed- ish speaking European countries. The review was conducted in a systematic way, following a review checklist consisting of 12 questions with multiple-choice answers. The checklist was partly based on concepts of content analysis methods by Krippendorff (1980).

Some of the issues that were addressed in the checklist concerned the understanding of the biodiversity concept, the type of impacts identified, the methods and tools used, the spatial and temporal scales studied, and the ecological levels that were considered in the assessment.

PROTECTION OF BIODIVERSITY

Since the Rio Summit and the ratification of the CBD (UNCED, 1992), biodiversity has received considerable attention at the international level. An integral part of the sustainability theory, the protection of biodiversity has emerged to become one of today’s environmental challenges. However, neither biodiversity nor sustainability are straightfor- ward concepts to understand. Sustainability is by itself a complex and somehow ambiguous concept (Edvarson, 2004). The work produced at the international, national and regional levels in the form of conferences, summits, conventions or policies gives wit- ness to the increasing awareness on biodiversity issues (Wegner et al., 2005). The term biodiversity itself is often misunderstood and lacks a clear definition (Wegner et al., 2005).

However, a definition of biodiversity, often referred to in the EIA context, is the one proposed in Article 2 of the CBD: "Biologi- cal diversity means 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” (UNCED, 1992; OJ, 1993). The definition recognises three different levels of biodiversity, the genetic, species and ecosystem levels. Other authors also include a landscape level of biodiversity (CEQ, 1993;

Geneletti, 2002; Noon and Dale, 2002), and Forman and Godron (1986) point out that a landscape is a collection of ecosystems.

Figure 1. Length of the road and railway projects

3

(13)

While discussing the different levels of biodiversity, confusion can occur concerning the use of the terms level and scale. According to Wiens (2002), scale tends to be associated with a physical dimension for the description of an area (e.g. local or regional scales) as opposed to level, which is associated with the spatial description of ecological processes (e.g. ecosystem or landscape levels). Finally, he complexity of the concept of biodiversity may pose a problem for its consideration in the EIA process.

At the national level many countries have strengthened their actions on the conservation of biodiversity. In Sweden, 15 environmental quality objectives were adopted by Parliament in 1999 (Gov. Bill 1997/98:145).

Among the objectives, many of them include the consideration of biodiversity (e.g. Flour- ishing Lakes and Streams; Thriving Wetlands;

Sustainable Forests; and A Varied Agricul- tural Landscape). More recently, an objective dealing specifically with biodiversity has been discussed and will be proposed to the Parlia- ment by the end of 2005. The implementation of the environmental quality objectives could help to translate into practice some of the ideas and recommendations presented in the CBD. One of the five fundamental principles for the environmental quality objectives is the preservation of biological diversity (Edvarsson, 2004). However, these rather vague principles (e.g. promotion of human health or wise management of natural resources) are integrated in the 15 environmental quality objectives that translate their content into more practical and specific goals. A challenge remains to establish a link between the environmental quality objectives and physical planning. Limiting the scope of this thesis to the planning sector, it is apparent that there is a variety of tools available to assist government agencies, institutions and others in reaching the objectives. The EIA and more recently the SEA processes are highly relevant to operationalise the intent expressed in the environmental objectives.

THREATS TO BIODIVERSITY

The need for protection of biodiversity stems from the diverse threats affecting biodiversity

worldwide and the emerging consciousness about the ongoing loss of biodiversity.

Among these threats, habitat loss and fragmentation are of major concern to the world’s biodiversity (Fahrig, 1997; Wilcove et al., 1998). In Sweden, an assessment of red- listed species presents a picture of frag- mented biotopes, such as primeval forest and unfertilised grassland, resulting in the isolation of populations and posing a major threat to their sustainability (Gärdenfors, 2000). The expansion of urban areas is threatening biodiversity on a global scale (Ricketts and Im- hoff, 2003). In the case of the Stockholm region, urban expansion is one of the main causes for the fragmentation of habitats. The development of housing, industries and related infrastructures steadily increases the pressure on the ecological assets of the region (Paper III). As part of or as a consequence of the urbanisation process, the development of infrastructure contributes significantly to problems of habitat loss and fragmentation (Forman, 2000; Trocmé et al., 2002; Paper III). There are numerous ecological impacts associated with transport infrastructures such as road and railways.

Fragmentation and barrier effects, in particular, are major impacts of linear projects (Trocmé et al., 2002). Together, habitat loss and isolation are causing fragmentation.

Habitat loss, which is one of the major impacts that occur during the construction phase, is caused by changes in the land use in the area affected by the project. Other impacts occur during the operation phase of the project. Examples of such impacts are noise and air pollution. In addition, disturbances can be related to recreational activities and the fact that access to previously remote areas has been facilitated. In the case of linear transportation projects, a number of impacts are linked to barrier effect and contribute to the isolation process. Barrier effects are caused by a combination of other impacts such as the physical hindrance caused by fences along roads, the changes in surface types, by the traffic itself (traffic mortality), the visual impact or the noise generated by the traffic (Seiler, 2002). Finally, other impacts can be associated with the decommis-

4

(14)

sioning phase of the project. One important characteristic that is common to many of these impacts is their spatial component, in the sense that they are widespread over large areas.

EIA AND SEA: LEGISLATIVE TOOLS FOR THE PROTECTION OF BIODIVERSITY

Different planning strategies or project de- signs can vary in density, distribution pattern, location or design and influence the nature of the impacts on biodiversity. An assessment and comparison of the different alternatives or strategies would provide relevant information to decision makers. In order to perform such assessments, specific planning tools have been adopted and implemented in the form of the EIA process and, more recently, the SEA process. These legislative tools aim to assess and limit impacts on the environment, including biodiversity. Impacts on the natural environment and biodiversity have been part of the scope of EIA since its incep- tion in the 1970’s, but the use of the term biodiversity in the EIA context is more recent. In Europe, the directive on the assessment of the effects of certain public and private projects on the environment or EIA directive (Official Journal of the European Communities, OJ, 1985), and its amendment in 1997 (OJ, 1997) specifies that impacts on flora and fauna need to be considered. More recently, the directive on the assessment of the effects of certain plans and programmes on the environment, or SEA directive (OJ, 2001), specifies that biodiversity as well as flora and fauna need to be part of the assessment, stressing the need for quality biodiversity assessment. During recent years, the importance of biodiversity in EIA and SEA have been pointed out and strengthened (CBD, 2004). In Europe, both the EIA and SEA directives are connected to the directive on the conservation of natural habitat and of wild fauna and flora (OJ, 1992), or Habitat directive. Should any policy, plan or pro- gramme affect any site listed under the Habi- tat directive, an SEA is required (Sheate et al., 2004). In the case of the EIA directive, such sites act as a selection but not automatic

criterion to determine if an EIA is required (Annex III, OJ, 1992). The EIA and SEA processes play different roles in the planning process and are applied at different levels in the planning scheme. In the case of a single project, an EIA can be performed whereas in the case of a policy, plan or program, a SEA would be the appropriate tool (Thérivel et al., 1992; Sadler and Verheem, 1996; Thérivel and Partidário, 1996; Hildén et al, 1998). The addition of the SEA process in recent years can be seen as a complement of the EIA process, ensuring that assessments are more proactive and occur earlier in the planning process (Glasson et al., 1999; Balfors and Schmidtbauer, 2002). Moreover, the SEA process allows and facilitates the consideration of large scale and cumulative impacts, which EIA on individual schemes face problems to address (Lee and Walsh, 1992; Glas- son et al., 1999).

The four countries included in the study:

Sweden, France, the UK and Ireland have different histories when it comes to their EIA legislation, but they all, as EU members countries, have to comply with the EIA directive (OJ, 1985). The temporal variation, the manner in which and reason for the implementation of EIA into the respective national legislations (e.g. nature protection, safety or for specific sectors of activity) has most probably influenced practices. More- over, specific traditions and regulations on nature protection might also imply dissimi- larities in the way the EIA process is per- ceived and applied in these countries.

EIA review

Even though all EISs deal with either road or railway projects, there was still a large hetero- geneity in the quality of the assessments of impacts on biodiversity. Firstly, the term biodiversity was not mentioned in 18 out of the 38 reports and, when present, was not defined or explained in any way. Figure 2 shows the frequency of use of the term biodiversity as well as the context in which it was used in the reports. According to the review, all reports contained a qualitative assessment whereas a quantification of potential impacts was performed in only eight 5

(15)

reports. As a consequence, many of the assessments were more descriptive in nature rather than analytical and predictive.

Figure 2. Uses of the term b odiversity in the EISs and frequencies of the different uses. i Five particular EISs from the UK, Ireland and Sweden contained a biodiversity assessment that was structured and presented in such a way that it took into account a number of stages. By stages it is meant that a distinction was made between impact magni- tude and significance (in the UK and Ireland) and between effects and consequences (in Sweden). A number of EISs made the rec- ommendation that different stages in the assessment should be considered, yet they failed to do so themselves.

A clear majority of EISs (90%) studied impacts during both construction and operation of the project (Figure 3). However, the information provided on impacts during construction was in most cases standardised and therefore not specific to the project. A distinction between short-term and long-term impacts was presented in five reports. Fur- ther, nine reports included information on follow-up studies or monitoring. It is worth noting here that the inclusion of a monitoring program is not a requirement according to the EU directive on EIA.

Figure 4 presents the terminology that was used in the biodiversity assessment section of

the reports. It provides information on the key ecological terms used in the biodiversity assessment and, indirectly, on the ecological levels underlying the assessment. Both the terms biotope and habitat were included in the review, even though their meanings are very similar, due to the fact that these terms were used in EISs from both Sweden and France. A habitat is defined by the space where a species live, including biotic and abiotic resources (Morrison and Hall, 2002) whereas a biotope is an area defined by specific physical, topographical, climatic and chemical parameters within which develops a specific ecological community (DIREN, 2002). A majority of reports addressed issues concerning species, habitats and biotopes whereas information on population or ecosystems was mostly omitted. Apart from one, all reports provided information on areas having some level of nature protection.

Moreover, 17 reports from all four countries made references to the Habitat directive and

the Natura 2000 network.

Figure 3. Consideration of issues related to the time scale in the assessment.

Considering the nature of the projects (roads and railways), effects specific to linear projects such as fragmentation or barrier effects (Piepers et al., 2002) could be expected. Fig- ure 5 presents the number of reports that assessed fragmentation and barrier effect related impacts.

6

(16)

7 Finally, another result of the review concerned the landscape character assessment.

While looking for information on biodiversity in the EISs, an ambiguity was apparent in that information that would be relevant for biodiversity assessment could be found in the section dealing with landscape studies. The connections between the landscape and biodiversity assessments were obvious in some reports (e.g. land cover data, and mitigation measures concerning vegetation). However, even though some information presented in the landscape assessment could have been relevant for the biodiversity assessment, none of the EISs reviewed had an integrated biodiversity and landscape assessment.

PREDICTION TOOLS FOR

BIODIVERSITY ASSESSMENT

The EIA and SEA processes form a set of legislative tools that help to determine and assess potential impacts on biodiversity.

Within these processes a biodiversity assessment needs to be performed using specific tools and methodologies. The EIS review performed in this study, as well as previous reviews (Treweek et al., 1993, Thompson et al., 1997; Byron et al., 2000; Atkinson et al., 2000) illustrates the lack of impact quantification and prediction in biodiversity assessments. Moreover, the spatial nature of many biodiversity related impacts induced by urbanisation and infrastructure developments strengthen the relevance of the use of GIS and spatial modelling tools in these assessments.

i i

Figure 4. Frequency of the use of key ecological terms in the b odivers ty assess- ment

Figure 5. Frequency of the consideration of impacts associated w th fragmenta ion ani t d barrier effects

The literature points out the potential use of GIS in the EIA process in general (e.g. João and Fonseca, 1996, Rodriguez-Bachiller, 2000) and in biodiversity assessment in particular (Treweek and Veitch, 1996; Geneletti, 2002). This potential use of GIS related tools ranges from basic display functions to ad- vanced modelling capabilities and the integration with existing models (Paper II). The tentative integration of ecological models and GIS technologies during the last 15 years (Goodchild, 2002) has meant that there are now potential applications available in the impact assessment field. More generally,

(17)

Clarke et al. (2000) stressed the role of the integration of GIS and environmental models to solve environmental problems. The following section presents a review of such GIS methods and GIS-based ecological models that offer potential for the assessment of impacts on biodiversity.

GIS in biodiversity assessment

The advantages of using GIS in ecological studies are presented in many EIA guidelines (e.g. CEQ, 1993; CEAA, 1996; World Bank, 1997), SEA guidelines (European Commis- sion, 1999) as well as in the literature (e.g.

Treweek and Veitch, 1996; João and Fonseca, 1996; Vanderhaegen and Muro, 2004; Paper II). However, the results from the EISs review (Paper I) illustrate the poor use of GIS to assess biodiversity related impacts and the fact that in most cases GIS is used primarily for its display capabilities and not for its analytical functionalities.

Depending on factors such as data availability, time spent on data processing and budget constraints, GIS could be used at different levels of complexity. GIS can be used to gather and build a spatial database (Burrough and McDonnell, 2000). This would allow and facilitate the integration of spatial data from different sources (Geneletti, 2002). The European Environment Agency used a spatial database composed of topography (digital elevation model), land cover, designated sites (under international conventions), nature inventories, population density, ecological regions, water pattern, administrative boundaries and coastal zones to perform a SEA on the trans-European transport network (European Environment Agency, 1998). This database was limited to available data at the European level and the report points out limitations due to data heterogene- ity. Other relevant digital data that could have been included were a soil map (geology), climatic variables (e.g. precipitation and tem- perature) and information on the hydrology and hydrogeology.

Additional information collected during field surveys (inventories, delineation, etc.), if conducted using a GPS (global positioning system), could be integrated into the existing

database. Data collection and generation of the database are essential in drawing up the baseline for the assessment. This database could be used for the production and presentation of maps (João and Fonseca, 1996) and thereby fulfil the visualisation and communi- cation objectives of the EIA and SEA processes. Krisp (2004) used the development of three-dimensional techniques in GIS to visu- alise ecological barriers resulting in habitat fragmentation in the case of urbanisation projects.

GIS could be used as a platform to apply different methodologies (Geneletti, 2002) and to perform basic spatial operations such as of buffer zones to delineate areas affected by disturbances (e.g. noise along roads). One of the most relevant data layers when dealing with biodiversity assessment is probably land- cover. Land-cover data are available at different spatial resolutions and can vary in the number of classes mapped and ecological information displayed. Moreover, the use of the original satellite images or aerial photographs can provide further information on land cover, for instance concerning vegetation and other useful ecological information.

For example, Löfvenhaft et al. (2002) used infrared aerial photographs to produce a biotope map of the greater Stockholm area.

Treweek and Veitch (1996) performed an assessment based on land cover categories and their proximity to the planned development in order to compare potential impacts of different project alternatives. Another method that has been used to compare different project alternatives was proposed by Geneletti (2003) and was based on ecosystem rarity.

GIS-based ecological models

The benefits of integrating GIS technologies and ecological modelling have been recog- nised by for example Goodchild (2002).

Within the research arena, GIS-based ecological modelling has been a growing branch of conservation biology and spatial ecology over the last decade (e.g. Guisan and Zim- merman, 2000; Lehmann et al., 2002).

8

(18)

9

Table 1 . A selection of ecological models and their characteristics

(19)

There are different ways to classify ecological models. A classification could be based on the modelling techniques and statistical methods used (Paper I) or could take into account the aim or potential use of the model (Decoursey, 1992). Bearing in mind the potential sectors for which each model was designed, one could also use data requirements and the limitations it induces as a parameter to distinguish between models.

The terminology used to describe ecological models can be somewhat confusing, but a number of key concept and terms associated to ecological models are easily identified.

However, no single classification will render a clear picture of the diversity of existing ecological models. The diversity in modelling approaches also allows for a variety of com- binations between models. Table 1 presents a non-exhaustive list of ecological models, including some of their characteristics. There are no clear distinctions between the different types of models as many of them integrate different modelling theories and techniques.

A number of key concepts and terms, however, are used to attempt to characterise and differentiate them.

Predictive modelling is used to describe a wide range of models, allowing predictions of, for example, species occurrences, population viability, or the location of suitable habitat.

An expert model relies on information gathered in literature or through the help of experts. In practice, a number of parameters implemented in the model are directly derived from existing information and not from new empirical data (e.g. LEDESS, Knol et al., 1999; HSI, Hays et al., 1981; LARCH, Ver- boom et al., 2001). An empirical model derives parameters from empirical data that are ana- lysed statistically or from using machine learning methods. If empirical, statistical and phenomenol- ogical models have many similarities (Guisan and Zimmermann, 2000), a number of nu- ances do however exist. Empirical models are used to describe the distribution of a de- pendent variable derived from empirical data (such as species occurrences), from relationships and correlations with environmental variables or predictors. Phenomenological mod- els aim to establish links such as correlation

between variables or probabilities of occur- rences. According to Maurer (2002) phenome- nological models tend mainly to describe the relationships between habitats and popula- tions. A statistical model corresponds to a model in which the parameters are derived from the statistical analysis of field data (e.g.

GRASP, Lehmann et al. 2002, 2003). Both statistical and machine learning models (e.g.

MAXENT, Phillips et al., 2004, GARP, Stockwell and Peters, 1999) are based on the study of empirical data. However, even though machine-learning models (e.g. including neural networks, Lusk et al., 2002) are em- pirically based, they use alternative algo- rithms, ordination and correlation techniques, which can change and adapt to new data, instead of the traditional statistical techniques such as regression analyses.

Some of the modelling techniques, such as regression methods, are based on data on presence (or abundance) and absence of biodiversity components (e.g. GRASP), whereas other techniques can model presence-only data (BIOCLIM, Busby, 1991;

GARP, Stockwell and Peters, 1999;

BIOMAPPER, Hirzel et al., 2002). Mechanistic models aim to describe detailed relationships, often at the level of individuals (Maurer, 2002), and are based on cause/effect relationships instead of a probability of occur- rence in the case of phenomenological models.

Metapopulation models (e.g. METAPHOR, Verboom et al., 2001; RAMAS, Akcakaya, 2001) focus on population dynamics and viability. They can relate habitat to population processes such as colonisation and extinction. Further, parameters needed for long-term persistence of species can be esti- mated (Akcacaya, 2001; Opdam et al., 2002).

The distinctions between modelling techniques is however not easy and some models combine to a certain extent expert knowledge and empirical studies (e.g. RAMAS, METAPHOR).

DISCUSSION AND CONCLUSION

The review of EISs confirmed a number of shortcomings in the way biodiversity assessments are performed. Firstly, the concept of biodiversity, even though widely promoted in 10

(20)

the context of the EIA process (Slootweg and Kolhoff, 2003), was not evident in most reports. One reason for this might be a natural delay between recent recommendations and their application. It could also be related to the complexity inherent to the biodiversity concept, or to the lack of adequate methodologies and competences to fulfil the recommendations on biodiversity. Of concern was the fact that the majority of reports contained mainly qualitative biodiversity assessments without sufficiently quantifying and predicting impacts. The result was descriptive assessments with very little analysis present.

Previous studies had also pointed out the vagueness of the assessments (Treweek et al., 1993) and the lack of predictions (Byron et al., 2000).

Information on the time perspective attached to impacts is required according to Annexure III of the EU directive on EIA (OJ, 1997). In practice, the distinction between long-term and short-term impacts was however seldom made. More positive was the practice of making a distinction between impacts occur- ring during the construction and operation phases of the project, even though the pre- dicted impacts during construction were often standardised and not project specific. A lack of quantitative assessment could be seen as an obstacle to the prediction of long-term impacts. The spatial scale of the assessment remained confined to small-scale and protected areas. The fact that many of the assessments were performed at the habitat level explain why they failed to predict large-scale and widespread impacts. It also accounted for difficulties experienced in assessing the potential cumulative impacts. Finally, the review revealed that impacts typically associated with roads and railways, such as fragmentation and barrier effects, were not sys- tematically taken into consideration.

The potential offered by GIS methods, though not currently extensively used in practice for biodiversity assessments, is rec- ognised and rapidly developing. Using GIS- based ecological models, however, remains confined to the research field. If GIS-based ecological models were used for making predictions, the spatial resolution of the

assessment and in some cases the temporal resolution would be improved. One challenge faced when implementing GIS tools or GIS- based ecological models is to identify which model(s) or modelling technique(s) would be the most appropriate to fulfil the goals of the biodiversity assessment. This choice is influenced by a number of criteria. Firstly, the data availability (in a digital format) is probably the most critical. Data availability could be affected the physical access to data (widespread databases between different admini- strations), the rights to use the data or its secrecy (e.g. endangered species), the cost of the data and its non-existence. According to Elith (2002), data availability and quality are determining factors for the choice of the modelling technique to be applied in specific circumstances. The nature of the study area (terrestrial, aquatic) and the main interest of the area (unique vegetation, species of particular protection or social status, etc.) can also influence that choice. Further criteria to be considered include the original aim of the model when developed, the interpretation of the results provided by the model, and its adaptability. Most ecological models are often designed for research purposes and not for use in real- life projects (Zabel et al., 2002).

However, Lehmann et al. (2002) point out that ecological models have reached a level where they can be used as planning and assessment tools Finally, the financial aspect and cost for the implementation of the methods should not be neglected.

Through the study of current EIA practices, guidelines on biodiversity assessment as well as the potential offered by GIS-based tools and ecological models different approaches to biodiversity assessment can be identified.

Within the EIA process, there is a tendency to focus the assessment on patches with a recognized biodiversity value, while omitting to consider a broader spatial perspective.

This approach could be seen as a patchwork approach reflecting the assessments performed on isolated areas or patches, without considering the bigger picture. Further, the ecosystem approach has been promoted internationally in the EIA context (CEQ, 1993; CBD, 2004) as a functional approach at 11

(21)

the ecosystem level. It illustrates the intention to consider interactions and functions within the impacted ecosystem. In view of the opportunities offered by GIS-based ecological modelling, a habitat suitability approach could be put forward. This approach pro- poses the implementation of GIS-based ecological models as tools to perform biodiversity assessment in the EIA and SEA processes. It could act as a link between the patchwork and ecosystem approaches (Figure 6) allowing the concepts of the ecosystem approach to be operationalised.

While making distinctions between different approaches to biodiversity assessment, it is relevant to discuss what is considered to be an adequate scale at which to study the impacts on biodiversity. Current practices fa- vour the local scale, while assessments at larger physical scales or ecological levels are the exception. João (2002) pointed out the need to rethink the scale(s) used in the assessment and the consequences it could have on determining impact significance. It is possible to observe a greater interest for large-scale studies (e.g. the ecosystem approach or the research in landscape ecology), as these are more relevant for some of the ecological processes involved. While discussing the optimal scale(s) of biodiversity assessment, the lack of integration between landscape and biodiversity assessments could

also be a hindrance to large-scale assessments. There are obvious connections between biodiversity and landscape assessment, which might lead to ambiguities concerning the definition of their scopes. At the same time, it is apparent that some guidelines on landscape assessment include ecological issues, which are primordial components of the aesthetic value of landscapes (Swanwick, 2002).

In light of the lack of integration between biodiversity assessment and landscape assessment, it could be possible to suggest a

fourth approach to biodiversity assessment called landscape ecological approach. A character- istic of this approach would be its aim of integrating landscape considerations in the assessment of impacts on biodiversity. Land- scape ecology practitioners and researchers tend to integrate issues on aesthetical or cultural vales of the landscape in the scope of landscape ecology (Tress et al., 2005). This could act as an important step on the way to achieving integration within the assessment.

The implementation of selected principles from the landscape ecology research field (Gutzwiller, 2002; Wiens, 2002) and the use of GIS tools, GIS-based ecological models could help to address scale issues for biodiversity assessments (Figure 7).

Figure 6. Different approaches to biodiversity assessment and their relationship to assorted physical scales and ecological levels

12

(22)

Figure 7. Schematic representation of the integration of landscape ecology using GIS methods and GIS-based ecological models

Bridging the gap between ecology and physical planning (Opdam et al., 2002) is a key issue to protect and maintain biodiversity in an urbanising environment. Urbanisation and infrastructure developments lead to fragmentation of natural habitats, ultimately converg- ing through the creation of non-viable islands of nature in a long-term perspective. More- over, the study confirms the tendency within the EIA process to focus on impacts at the local scale, most often on formally protected areas, and the lack of concern for large scale, widespread or cumulative impacts. The need for prediction tools is self-evident. GIS-based ecological models, in particular, offer possibilities for making predictions of widespread impacts and taking into account ecological processes. The use of such tools would, in turn, provide decision makers with relevant information and data on the comparison of different alternatives or scenarios. However, the scale at which to undertake the biodiversity assessment can be problematic. An absence of integration between biodiversity and landscape assessments accentuates the problems in impact prediction and assessment at larger scales. Ultimately, the comparison of different alternatives and scenarios based on quantitative and spatially distributed predictions of impacts on biodiversity would contribute to a more sustainable planning system.

FUTURE RESEARCH

The need for methods to improve adequate predictions of impacts from urbanisation and infrastructure development on biodiversity has been established. However, a number of issues have been identified that need further exploration. Even though GIS methods and GIS-based ecological models offer great potential for predictions, it remains to ex- plore these possibilities and test prediction methods and tools to ensure their applicability in physical planning. Data requirements and the need for expert knowledge are two parameters that might limit their applicability.

A comprehensive evaluation and comparison between the different methods could help to clarify their limitations and the extent of their applicability.

Another issue that needs attention is the physical scale and ecological level that should be considered in biodiversity assessment and their relevance for the EIA and SEA processes. In fact, the geographical scale or ecological level, inherent to predictive models might influence their potential implementation depending on the size of the development under consideration.

Finally, the interpretability and relevance of the obtained predictions need to be discussed in the light of the goals aimed at the protection of biodiversity and sustainable development.

13

(23)

REFERENCE LIST

Akçakaya, H.R., 2001. Linking population-level risk assessment with landscape and habitat models. The Science of the Total Environment, 274:283-291.

Atkinson, S.F., Bhatia, S., Schoolmaster, F.A., Waller, W.T., 2000. Treatment of biodiversity impacts in a sample of US environmental impact statements. Impact Assessment and Project Appraisal, 18:271-282.

Balfors, B., Schmidtbauer, J., 2002. Swedish Guidelines for Strategic Environmental Assessment for EU Structural Funds.

In: European Environment, 12:35-48.

Burrough, P.A., McDonnell, R.A., 2000. Principles of Geographical Information Systems. Oxford University Press, New York, 333 pp.

Busby, J.R., 1991. BIOCLIM - A Bioclimatic Analysis and Prediction System. In: Margules, C.R.& M.P. Austin (eds.) Nature Conservation: Cost Effective Biological Surveys and Data Analysis, Canberra: CSIRO, 64-68.

Byron, H.J., Treweek, J.R., Sheate, W.R., Thompson, S., 2000. Road developments in the UK: An Analysis of Ecological Assessment in Environmental Impact Statements Produced between 1993 and 1997. Journal of Environmental Planning and Management, 43:71-97.

Clarke, K., Parks, B., Crane, M., 2000. Preface: A perspective on GIS-environmental model integration (GIS/EM). Journal of Environmental Management, 59:229-233.

Canadian Environmental Assessment Agency (CEAA), 1996. A Guide on Biodiversity and Environmental Assessment.

Ministry of Supply and Services, Canada, 15 pp.

Council on Environmental Quality (CEQ), 1993. Incorporating Biodiversity Considerations into Environmental Impact Analysis under the National Environmental Policy Act (CEQ, Washington DC), 29 pp.

Decoursey, D.G., 1992. Developing models with more detail: do more algorithm give more truth? Weed Technology, 6:109- 715.

Direction régionale de l’environnement de Midi-Pyrénées, 2002. Guide sur la prise en compte des milieux naturels dans les études d’impact, 75 pp.

Edvarsson, K., 2004. Using Goals in Environmental Management: The Swedish System of Environmental Objectives. Envi- ronmental Management, 34:170-180.

Elith, J., 2002. Quantitative Methods for Modeling Species Habitat : Comparative Performance and an Application to Austra- lian Plants. In: S., Ferson, M., Burgman, (eds.) Quantitative Methods for Conservation Biology. New York, Springer- Verlag, 39-58.

European Commission, 1999. Manual on Strategic Environmental Assessment of Transport Infrastructure Plans. Directorate- General VII, Brussels, 133 pp.

European Environment Agency, 1998. Spatial and Ecological Assessment of the TEN: Demonstration of Indicators and GIS methods. Environmental Issues Series 11, EEA, Copenhagen, 52 pp.

Fahrig, L., 1997. Relative effects of habitat loss and fragmentation on population extinction. Journal of Wildlife Manage- ment, 61:603-610.

Forman, R.T.T., 2000. Estimate of the area affected ecologically by the road system in the United States. Conservation Biology, 14:31-35.

Forman, R.T.T., Godron, M., 1986. Landscape Ecology. John Wiley and Sons, New York, 619 pp.

Geneletti, D., 2002. Ecological evaluation for environmental impact assessment. Netherlands Geographical Studies, (NGS 301), 218 pp.

Geneletti, D., 2003. Biodiversity Impact Assessment of roads: an approach based on ecosystem rarity. Environmental Impact Assessment Review, 23:343–365.

Glasson, J., Therivel, R., Chadwick, A., 1999. Introduction to Environmental Impact Assessment. Principles and procedures, process, practice and prospects. 2^nd edition. UCL Press: London, 496 pp.

Goodchild, M.F., 2002. Preface. In: J.M. Scott, P.J. Heglund, M.L. Morrison, J.B. Haufler, M.G. Raphael, W.A. Wall, Samson, F.B., (eds.) Predicting species occurrences: Issues of accuracy and scale. Washington: Island Press, xv-xvii.

Gov. Bill 1997/98:145. Svenska miljömål. Miljöpolitik för ett hållbart Sverige.

Guisan, A., Zimmermann, N.E., 2000. Predictive habitat distribution models in ecology. Ecological modelling, 135:147-186.

Gutzwiller, K.J., 2002. Applying Landscape Ecology in Biological Conservation: Principles, Constrains, and Prospects. In:

K.J., Gutzwiller, (Ed.) Applying Landscape Ecology in Biological Conservation. New York: Springer, 481-495.

Gärdenfors, U., (ed.) 2000. Rödlistade arter i Sverige 2000 – The 2000 Red List of Swedish Species. ArtDatabaken, SLU, Uppsala, 397 pp.

14

(24)

Hays, R.L., Summers, C., Seitz, W., 1981. Estimating wildlife habitat variables. USDI Fish and Wildlife Service. FWS/OBS- 81/47, 111 pp.

Hildén, M., Valve, H., Jónsdóttir, S., Balfors, B., Faith-Ell, C., Moen, B., Peuhkuri, T., Schmidtbauer, J., Swensen, I., Tesli, A., 1998. EIA and its application for policies, plans and programmes in Sweden, Finland, Iceland and Norway. Te- maNord 1998:567. Nordic Council of Ministers, Copenhagen, 169 pp.

Hirzel, A.H., Hausser, J., Chessel, D., Perrin, N., 2002. Ecological-niche factor analysis: How to compute habitat- suitability maps without absence data? Ecology, 83:2027-2036.

João, E., 2002. How scales affects environmental impact assessment. Environmental Impact Assessment Review, 22:289- 310.

João, E., Fonseca, A., 1996. The Role of GIS in Improving Environmental Assessment Effectiveness: Theory vs Practice.

Impact Assessment, 14(4):371-387.

Knol, W.C., van Eupen, M., Roos-Klein-Lankhorst, J., Verweij, P., 1999. Landscape Ecological and Evaluation Support System LEDESS. GIS Application Manual version 2.0. Winand Staring Centre Wageningen. Technical Manual, 112 pp.

Krippendorff, K., 1980. Content Analysis. Sage Publication, Beverly Hills, Calif., 191 pp.

Krisp, J. M., 2004. Three-dimensional visualisation of ecological barriers. Applied Geography, 24:23-34.

Lee, N., Walsh, F., 1992. Strategic environmental assessment: an overview. Project appraisal, 7(3):126-136.

Lehmann, A., Overton, J.McC., Austin, M.P., 2002. Regression models for spatial prediction: their role for biodiversity and conservation. Biodiversity and Conservation, 11:2085-2092.

Lehmann, A., Overton, J.McC., Leathwick, J.R., 2003. GRASP: generalized regression analysis and spatial prediction.

Ecological Modelling, 160:165-183.

Lusk, J.J., Guthery, F.S., DeMaso, S.J., 2002. A Neural Network Model for Predicting Northern Bodwhite Abundance in the Rolling Red Plains of Oklahoma. In: J.M. Scott, P.J. Heglund, M.L. Morrison, J.B. Haufler, M.G. Raphael, W.A. Wall, F.B. Samson, (eds.) Predicting species occurrences: Issues of accuracy and scale. Washington: Island Press, 345-355.

Löfvenhaft, K., Björn, C., Ihse M., 2002. Biotope patterns in urban areas: a conceptual model integrating biodiversity issues in spatial planning. Landscape and Urban Planning, 58:223-240.

Maurer, B.A., 2002. Predicting Distribution and Abundance: Thinking within and between scales. In: J.M. Scott, P.J. He- glund, M.L. Morrison, J.B. Haufler, M.G. Raphael, W.A. Wall, F.B. Samson (eds.) Predicting species occurrences: Is- sues of accuracy and scale. Washington: Island Press, 125-132.

Morrison, M.L., Hall, L.S., 2002. Standard Terminology: Toward a Common Language to Advance Ecological Understand- ing and Application. In: J.M. Scott, P.J. Heglund, M.L. Morrison, J.B. Haufler, M.G. Raphael, W.A. Wall, F.B. Samson (eds.) Predicting species occurrences: Issues of accuracy and scale. Washington: Island Press, 43-52.

Noon, B.R., Dale, V.H., 2002. Broad-Scale Ecological Scinece and Its Application. In: K.J., Gutzwiller, (Ed.) Applying Landscape Ecology in Biological Conservation. New York: Springer, 34-52.

Official Journal of the European Communities, 1985. Council Directive 1985/337/EEC of 27 June 1985 on the assessment of the effects of certain public and private projects on the environment, Official Journal L175, 05/07/1985.

Official Journal of the European Communities, 1992. Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora. Official Journal L 206, 22/07/1992.

Official Journal of the European Communities, 1993. Convention on Biological Diversity. Official Journal, L 309/1, 13/12/1993.

Official Journal of the European Communities, 1997. Council Directive 97/11/EC of 3 March 1997, amending Directive 85/337/EEC on the assessment of effects of certain public and private projects on the environment. Official Journal L 073, 14/03/1997.

Official Journal of European Communities, 2001. Directive 2001/42/EC of the European Parliament and the Council of the 27 June on the assessment of the effects of certain plans and programmes on the environment. Official Journal L 197/30, 21/07/2001.

Opdam, P., Foppen, R., Vos, C., 2002. Bridging the gap between ecology and spatial planning in landscape Ecology. Land- scape Ecology, 16:767-779.

Phillips, S.J., Dudík, M., Schapire, R.E., 2004. A Maximum Entropy Approach to Species Distribution Modeling. Proceed- ings of the 21^st International Conference on Machine Learning, Banff, Canada.

Piepers, A., Alvarez, G., Bouwma, I.M., De Vries, H.J.G., Seiler, A., 2002. Minimizing Fragmentation through Appropriate Planning. In: M., Trocmé, S., Cahill, J.G., de Vries, H., Farrall, L., Folkeson, G., Fry, C., Hichs, J., Peymen, (eds.) COST 341. Habitat Fragmentation due to Transportation Infrastructure. The European Review. Office for Official Publi- cations of the European Communities. Luxemburg, 115-128.

Ricketts, T., Imhoff, M., 2003. Biodiversity, urban areas, and agriculture: locating priority ecoregions for conservation.

Conservation Ecology, 8:1.

15

(25)

Rodriguez-Bachiller, A., 2000. Geographical Information Systems and Expert Systems for Impact Assessment. Part I : GIS.

Journal of Environmental Assessment Policy and Management, 2(3): 369-414.

Sadler, B., Verheem, R., 1996. Strategic Environmental Assessment: Status, Challenges and Future Decisions, Document 53.

Ministry of Housing, Spatial Planning and Environment, The Hague, 188 pp.

Seiler, A., 2002. Effects of Infrastructure on Nature. In: M., Trocmé, S., Cahill, J.G., de Vries, H., Farrall, L., Folkeson, G., Fry, C., Hichs, J., Peymen, (eds.) COST 341. Habitat Fragmentation due to Transportation Infrastructure. The European Review. Office for Official Publications of the European Communities. Luxemburg, 31-50.

Sheate, W.R., Byron, H.J., Smith, S.P., 2004. Implementing the SEA Directive: sectorial challenges and opportunities for the UK and EU. European Environment, 14:73-93.

Slootweg, R., Kolhoff A., 2003. A generic approach to integrate biodiversity considerations in screening and scooping for EIA. Environmental Impact Assessment Review, 23:657-681.

Stockwell, D.R.B., Peters D.P., 1999. The GARP modelling system: Problems and solutions to automated spatial prediction.

International Journal of Geographic Information Systems, 13:143-158.

Swanwick, C., 2002. Landscape Character Assessment – Guidance for England and Scotland. Countryside Agency, Chelten- ham and Scottish Natural Heritage, Battleby. 84 pp.

Therivel, R., Wilson, E., Thompson, S., Heaney, D., Pritchard, D., 1992. Strategic Environmental Assessment. Earthscan:

London, 158 pp.

Therivel, R., Partidário, M. R., 1996. The Practice of Strategic Environmental Assessment. Earthascan: London, 206 pp.

Thompson, S., Treweek, J.R., Thurling, D.J., 1997. The ecological component of environmental impact assessment: a critical review of British environmental statements. Journal of Environmental Planning and Management, 40(2):157-171.

Tress, B., Tress, G., Fry, G., Opdam, P., (eds.) 2005. From landscape research to landscape planning: Aspects of integration, education and application. Springer. (In press).

Treweek, J., Veitch, N., 1996. The potential application of GIS and remotely sensed data to the ecological assessment of proposed new road schemes. Global Ecology and Biogeography Letters, 5:249-257.

Treweek, J.R., Thompson, S., Veitch, N., Japp, C., 1993. Ecological assessment of proposed road developments: a review of environmental statements. Journal of Environmental Planning and Management, 36:295-307.

Trocmé, M., Cahill, S., de Vries, J.G., Farrall, H., Folkeson, L., Fry. G., Hichs, C., Peymen, J., (eds.) 2002. COST 341.

Habitat Fragmentation due to Transportation Infrastructure. The European Review. Office for Official Publications of the European Communities. Luxemburg, 253 pp.

United Nations Conference on the Environment and development (UNCED), 1992. The UN Convention on Biological Diversity (Rio, UNCED).

United Nations Educational and Cultural Organizations (UNESO), 1971. Convention on Wetlands of International Impor- tance especially as Waterfowl Habitat. Ramsar on 2 February, 1971. The convention on Wetlands text, as amended in 1982 and 1987.

Vanderhaegen, M., Muro, E., 2004. Contribution of a European spatial data infrastructure to the effectiveness of EIA and SEA studies. Environmental Impact Assessment Review, 25:123-142.

Verboom, J., Foppen, R., Chardon, P., Opdam, P.F.M., Luttikhuizen, P., 2001. Introducing the key patch approach for habitat networks with persistent populations: an example for marshland birds, Biological Conservation, 100:89-101.

Wegner, A., Moore, S.A., Bailey, J., 2005. Consideration of biodiversity in environmental impact assessment in Western Australia: practitioners perceptions. Environmental Impact Assessment Review, 25:143-162.

Wiens, J.A., 2002. Central Concepts and Issues of Landscape Ecology. In: K.J., Gutzwiller, (Ed.) Applying Landscape Ecology in Biological Conservation. New York: Springer, 1-21.

Wilcove, D.S., Rothstein, D., Dubow, J., Phillips, A., Losos, E., 1998. Quantifying threats to imperilled species in the United States. BioScience, 48:607-615.

World Bank, 1997. Environmental Assessment sourcebook Update: Biodiversity and Environmental Assessment. The World Bank, Washington DC, 10 pp.

Zabel, C.J., Roberts, L.M., Mulder, B.S., Stauffer, H.B., Dunk, J.R., Wolcott, K., Solis, D., Gertsch, M., Woodbridge, B., Wright, A., Goldsmith, G., Keckler, C., 2002. A Collaborative Approach in Adaptive Management at a Large-landscape Scale. In: J.M. Scott, P.J. Heglund, M.L. Morrison, J.B. Haufler, M.G. Raphael, W.A. Wall, F.B. Samson (eds.) Predict- ing species occurrences: Issues of accuracy and scale. Washington: Island Press, 241-253.

OTHER REFERENCES

Convention on Biological Diversity, 2004. Decision VII/11 Ecosystem approach, URL:

http://www.biodiv.org/decisions/default.aspx?m=COP-07&id=7748&lg=0. Accessed in March 2005.

16