Assessment of Biodiversity at the Local Scale for Environmental Impact Assessment and Land-use Planning

Abstract

This paper examines approaches to biodiversity assessment for environmental impact assessment in the context of an increasing emphasis on the conservation of biodiversity beyond the confines of protected sites, and the role of ecosystem services in sustainable development. The paper argues that while the matrix approach adopted in the UK for environmental impact assessment has some value in providing a consistent national framework, for ecological impact assessments it lacks sensitivity and focuses attention on notified sites and species. It therefore under-values those ecological resources at local scales. Three case studies are briefly considered where evaluation employed local assessment criteria, including an appraisal of the impacts on ecosystem integrity and the functioning of ecological networks. Attention is drawn to the importance of national and local biodiversity action plan habitats and species where trend data provide one of the key criteria for inclusion in the action plans and the England Biodiversity List. Evaluations based on the presence of such habitats and species need to take ecological and geographical context into account.

Introduction

In the UK, approaches to the setting of criteria for the assessment of the nature conservation value of a defined area of land commenced with the A Nature Conservation Review (Ratcliffe, 1977). A range of qualities such as naturalness, diversity, rarity, fragility and size provide a framework for the selection of national sites for statutory protection. Since then, the development of guidelines for the selection of the UK's sites of special scientific interest (SSSI) have provided a set of more objective and defensible criteria by which such sites of national value may be selected (Nature Conservancy Council [NCC], 1989). Similar criteria apply to the selection of sites that qualify for international designations such as Ramsar sites and the Natura 2000 sites in Europe, termed special protection areas and special areas of conservation in the UK (Stroud et al., 1990, 2001; Brown et al., 1997).

County wildlife trusts are active in managing local nature reserves and identifying sites that, while not meeting the criteria for designation as SSSI, are nevertheless of local or regional importance for nature conservation. Over the past decade, the trusts, sometimes in association with local planning authorities or biodiversity partnerships, have been addressing the need for a similarly robust approach to the definition and designation of sites of local conservation importance. These are then are identified in regional and local plans (The Wildlife Trusts, 1997) and the evolving regional spatial strategies and local development frameworks of the local planning authorities. While not statutorily protected, there is a presumption against developments that might adversely affect these sites.

In parallel to this process of evaluation of ecological interest for nature conservation objectives, the UK Department of Transport has developed ‘transport analysis guidance’ (TAG) for the assessment of ecological value as one of the sub disciplines of environmental impact assessment (EIA) for transport schemes in the New Approach to Appraisal (DETR, 1998).1

This appraisal methodology addresses the geographical scale at which the feature is important with reference to the national hierarchy of site designation as a definition of ecological value from the highest, international scale to undesignated sites deemed to be of some ecological interest at the local scale. This first stage of evaluation, as part of the ecological impact assessment approach recommended in TAG, is replicated in Table 1.

Table 1. TAG ecological evaluation

Criterion	Value	Examples
High importance and rarity on an international scale with limited potential for substitution	Very high	Internationally designated sites
High importance and rarity on a national or regional scale with limited potential for substitution	High	Nationally designated sites
	Regionally important sites with limited potential for substitution
High or medium importance and rarity at a local or regional scale with limited potential for substitution	Medium	Regionally important sites with potential for substitution.
	Locally designated sites
Low or medium importance and rarity on a local scale	Low	Undesignated sites of some local biodiversity and earth-heritage interest
Very low importance and rarity on a local scale	Negligible	Other sites with little or no local biodiversity and earth heritage interest

The value of the site is imported into another matrix that correlates this with the magnitude of impact to produce an overall appraisal of impact on an 8-point scale from ‘very large adverse’ impacts to ‘large beneficial’. The prime objective of this scale of evaluation is primarily to determine the acceptability of the development proposals and to determine which ecological receptors are in need of protection by appropriate mitigation and, implicit in the latter process, to develop a scale of priority for mitigation.

While this approach is useful in providing a national standard for ecological evaluation and impact assessment, it tends to focus attention on notified (both statutory and non-statutory) sites. A rigid application of this scheme has a number of disadvantages, some of which are considered below.

•	It is erroneous to conclude that all areas with an ecological interest at the national scale, or even at the international scale, have been identified or scheduled. The SSSI system in the UK originally aimed to cover only a representative selection of the best sites in the area of search or scheduling unit, not necessarily to notify all sites that could qualify.
•	Similarly not all areas of SSSI quality have necessarily been notified or identified as second-tier county wildlife sites: the coverage varies from county to county.
•	The dynamic nature of natural systems, changing land management, and continuing pressures on wildlife may result in intrinsic changes in ecological value or create new priorities for the conservation of biodiversity, particularly in the face of climate change.
•	The approach to the evaluation of species present outside notified (statutory and non-statutory) sites is not clear from the TAG approach.
•	Determination of ecological value in sites with no formal designations (the shaded boxes in Table 1) remains largely a matter of professional judgement and hence subject to variation depending on the skills and interests of the assessor.
•	There is a lack of sensitivity in the matrix approach; for example, all degrees of impact (i.e. from total loss to minor negative impacts) upon sites categorized as of ‘low value’ result in ‘slight adverse’ impact scores, deemed not significant in planning terms and therefore rarely attracting strategies for mitigation.
•	The ranking of interests from high (international scale) to low (local scale) can therefore tend to undervalue certain resources that may be of greater value at a local level and/or that may make an essential contribution to biodiversity in a local or regional context.

In respect of the last point in particular, Genelitti (2003) refers to these shortcomings in EIA where only notified sites and species are considered. A number of reviews of EIA suggest that ecological resources outside protected sites tend to be inadequately assessed and evaluated (Byron et al., 2000). From the experience of the author in reviewing ecology assessments as part of EIA, this failure continues to be widespread. The result is that ecological resources outside notified sites, or species not cited for special protection by the EC Birds or Habitats Directives or national legislation, tend to be consigned to low categories of value. The magnitude of impact is then consequently assessed as slightly adverse, which implies acceptability in planning terms and generally requiring no mitigation.

This runs counter to the aims of the 1992 Rio Convention on Biodiversity, which emphasizes the importance of local protection of habitats and species and the conservation of genetic diversity in the broader landscape. The conservation of biodiversity at all geographical scales is now reflected in UK government policies on sustainability and planning; for example, in its Planning Policy Statement No. 9 (2005) where paragraph 1.vi states:

Where a planning decision would result in significant harm to biodiversity and geological interests which cannot be prevented or adequately mitigated against, appropriate compensation measures should be sought. If that significant harm cannot be prevented, adequately mitigated against, or compensated for, then planning permission should be refused.

The UK Countryside and Rights of Way Act 2000 (HMSO, 2000) (as amended by the Natural Environment and Rural Communities Act 2006; HMSO, 2006) requires all public authorities to ‘have regard to biodiversity as far as is consistent with the proper exercise of their functions’. In addition, the Planning and Compulsory Purchase Act 2004 requires that planning authorities address government policies (e.g. PPS 9) in their regional spatial strategies and in local development schemes.

The focus is therefore shifting from a site-based approach to the conservation of selected habitats and species, to the maintenance of biodiversity, functioning ecosystems and the intrinsic interactions and relationships with human populations (Vermeulen & Koziell, 2002), or to use the market terminology, ecosystem services (DEFRA, 2007a; Haines-Young & Potschin, 2007). The increasing realization of the importance of ecological networks for wildlife conservation, particularly in the face of climate change (Bright, 1997; Jongman & Pungetti, 2004), reinforces the need for the evaluation and conservation of biodiversity not only at the national or regional scales but at local levels where the presence of key habitat linkages and core areas is vital.

This emphasizes the need to accurately value biodiversity at the local scale both in the context of both nature conservation initiatives and for the purpose of ecological impact assessment as part of an EIA. The Guidelines for Ecological Impact Assessment in the United Kingdompublished by the Institute of Ecology and Environmental Management (IEEM, 2006) have made an attempt to move away from the matrix approach. The guidelines still uses the principle of valuing an ecological resource at a defined geographic scale but advocate that impacts are evaluated simply as significant or not significant for the geographic level at which the ecological resource is valued. Whether an impact is significant is based on whether the impact affects the integrity (see IEEM, 2006) of a defined site or species. Therefore, the total loss of a site of importance at the parish (local) level would be assessed as being significant at the parish level. This approach provides a more accurate assessment of the impacts than a conclusion of Slight Adverse and can provide more accurate guidance on the requirements for mitigation to maintain biodiversity at the local level.

However, the IEEM guidelines do not set out a structured method to value habitats at local levels. To provide this structure, it may be valid to make recourse to the Ratcliffe (1977) criteria, which provide a list of primary qualities: size, diversity, naturalness, rarity, fragility and typicalness. Secondary criteria are: recorded history, position in an ecological or geographical unit, potential value and intrinsic appeal. These require a measure of professional judgement, as advised in the IEEM assessment guidelines, to be applied to the evaluation process within this framework. At the same time, there may be, as in the evaluation criteria for designated sites, a need for more objectivity to be applied to support professional judgement to make the case for site safeguard or sustainable development.

The following sections briefly review established and some arising approaches to ecological evaluation, and consider some of the issues and questions arising for the assessment of biodiversity at the local scale for the purpose of ecological impact assessment and land-use planning.

National Criteria for Biodiversity in the UK

Statutory Sites

Since the establishment of the Ratcliffe (1977) criteria, which in some cases may be considered to be somewhat subjective, requiring a degree of professional knowledge and judgement for their application, the guidelines for the selection of biological SSSI (NCC, 1989) have established a number of additional empirical measures.

For some of the Ratcliffe criteria (e.g. rarity, diversity), empirical guideline measures can be applied as described by the NCC (1989) to evaluate sites for SSSI selection.2 These measures depend on an adequate national database of ecological information; for example, bird census data, habitat data from plant community surveys (phase 1 habitat surveys). The National Vegetation Classification (NVC) (Rodwell, 1991, et seq.), in providing a description of plant communities throughout the UK and a guide to the national pattern of community distribution, has enhanced the evaluation of habitats in relation to measures of naturalness, diversity, typicalness, rarity and conservation importance. The descriptive texts in the NVC imparts a considerable body of information on aspects of ecological processes in the plant communities, succession and community relationships, enabling judgements to be made on fragility and ecological potential.

The main measures used to qualify the Ratcliffe criteria (NCC, 1989) are indicated in Table 2.3

Statutory Species

The matrix approach to evaluation when applied to species requires decisions to be made as to where the species should fall into the hierarchy and as to how this evaluation relates to the site or area under examination. It is relatively straightforward to assign protected species, identified by EC Annexes in the Birds and Habitats Directives, to the top category (international). Species identified in the 1981 Wildlife & Countryside Act schedules are generally accorded the second highest level of national importance. However, the presence of, particularly mobile, protected species should not automatically confer international or national level status unless it is clear that, from an appropriate assessment, the species gains key resources from the site such that the site has importance for population status. Criteria for the evaluation of species populations (e.g. size, rarity) are considered in Table 2.

Table 2. The ‘Ratcliffe criteria’

Ratcliffe criterion	Qualification	Examples of national and international measures
Primary criteria
Size	Of a size appropriate to maintain its function; for example, sites for top predators may need to be extensive, species rich grasslands can be limited in area but stable, populations should be at above threshold levels for viability and resilience	Some sites assessed in terms of the size of population supported; for example, numbers of over-wintering wildfowl or waders, breeding seabird colonies, bat roosts
		The 1% level of national and global bird populations is used respectively for the notification of SSSI (UK) and globally important bird areas (IBA) or key biodiversity areas (KBA)
Diversity	Number of species present (generally the species richness of a defined group) in a site. Comparisons of diversity should apply to sites of the same community type	The UK NVC provides descriptions of characteristic plant communities used as a baseline for comparison
		Corine biotopes describe vegetation communities in Europe
		Diversity indices (e.g. Simpson, Shannon) can provide comparative assessments between sites
Naturalness	Habitats largely unmodified by human activity (e.g. salt marsh, blanket bog). Some anthropogenic habitats retaining characteristics of low disturbance may be included (e.g. heathlands, meadows)	NVC communities, Corine biotopes, indicator species, presence and proportion of native and non-native species, professional judgement, recorded history of the site
Rarity	Applied to habitats or species where areas are limited, population numbers low or the habitat or species limited in distribution	NVC and Corine biotopes can define rare habitat types
		Habitats <10,000 hectares in the UK considered rare
		UK species occurring in <15, 10 × 10 km squares of the national grid are considered nationally rare (i.e. <0.5% of the total); from 16–99, 10 km squares (nationally scarce (<3.5% of the national total 10 km squares)
		Red Data Book species. International Union for Conservation of Nature (IUCN) Red List
		Endemics may be rare at particular scales
Fragility	Habitats or species vulnerable to disturbance and loss because of small area, low population or reliance on a single key resource	Mainly professional judgement backed up by research data or precedent. Red Data or IUCN Red listing may imply fragility or vulnerability. Trend analysis for species or species assemblages
		Small local populations may have particular importance to meta- populations in maintaining gene flow
		Certain endemics may be vulnerable to the effects of competition from alien species
Typicalness	Reflects the need to include the more abundant habitats communities and species, those at the centre of their natural range	NVC or Corine biotope analysis for plant communities, community and species distribution data, professional judgement
Secondary criteria
Recorded history	Sites with a good record of study may contribute to scientific knowledge and cultural awareness
Potential value	Degraded sites with remnants of the original ecological formation that may be readily restored	Professional judgement, although NVC community analysis may assist in identifying sites
		Habitat modelling; for example, physical habitat simulation (PHABSIM) for river systems, Predictive System for Multimetrics (PSYM) for ponds
Position in an ecological unit	Notes the value of contiguity of different habitats (the antithesis of fragmentation) or the particular ecological qualities of seral stages or ecotones. Can include wildlife corridors or networks	Vegetation community analysis (e.g. NVC, Corine) may assist in the identification of related communities according to physical gradients, ordination analyses
Intrinsic appeal	Habitats, and habitats supporting species, with particular public appeal. Can include areas scheduled for wildlife and public amenity, green corridors.	Professional judgement and/or results of stakeholder surveys and consultation. Certain ‘flagship’ species with intrinsic appeal may also act as ‘umbrella’ or ‘keystone’ species where their protection requires conservation of specific habitats or key ecological communities

However, with the enactment in the UK of the Countryside and Rights of Way Act 2000 and the subsequent amendments in the Natural Environment and Rural Communities Act 2006, a large number of habitats and species in the UK Biodiversity Action Plan (BAP) now have some legislative significance—in as much as Section 40 of Natural Environment and Rural Communities Act 2006 placed a duty on all public bodies and statutory undertakers ‘to have regard’ to the purpose of conserving biodiversity as defined by the key BAP species or habitats listed in Section 41 of the Act.

Two questions immediately arise from this. Should habitats and species in the BAP now be accorded importance at regional or national scales? Hitherto, it has been common in EIA to consider BAP species of local importance only and rarely is mitigation proposed. Under Article 6(3) and 6(4) of the European Community's Habitats Directive, an ‘appropriate assessment’ is required where development may indirectly affect the integrity of European protected sites by affecting protected species in areas beyond the site boundary, with mitigation as needed in order to maintain a favourable population status. Should a similar rationale apply to BAP species in the wider countryside?

Regional and Local Criteria for Biodiversity

Non-statutory Sites and Species

The Ratcliffe criteria are independent of scale; that is, they can be applied at scales ranging from the national dimension to assessments at county, borough, parish or neighbourhood level. Thus they continue to stand as key principles for ecological evaluation and have been re-iterated by many UK wildlife trusts, local planning authorities or biodiversity partnerships as criteria for the selection of non-statutory local wildlife sites (see Table 3) that are notified by local planning authorities for protection through the planning system (variously termed County Wildlife Sites, Sites of Nature Conservation Interest, Sites of Interest for Nature Conservation). In urban areas, the Ratcliffe criterion of ‘intrinsic appeal’ may be included as one of a number of social criteria in considering the importance of core sites, wildlife corridors and their accessibility for the public in the local area.

Table 3. Examples of criteria for selection of non-statutory wildlife sites

Primary criterion	Examples of qualifying criteria
Size	Woodland habitats >1 hectare, heaths >0.5 hectares (Dorset Wildlife Trust, 2001)
	The regular occurrence of 1% or more of the county population (e.g. of bird species)
	With 0.5% or more of the National population of bird species. (Derbyshire Biodiversity, undated; Hertfordshire Habitat Survey, undated)
Diversity	Defined NVC communities (Cambridgeshire County, 2006; Gwent Wildlife Trust, 2004; Hertfordshire Habitat Survey, undated) deemed to be important at a county level or within the Natural Area (Derbyshire Biodiversity, undated)
	Semi natural plant communities as defined by listed indicator species; for example, for ancient woodlands, grasslands, ponds (Gwent Wildlife Trust, 2004; Nottinghamshire Biological Record Centre, undated; Derbyshire Biodiversity, undated; Essex Wildlife Trust, 2007)
	Use of the Hedgerow Regulations 1997 to provide criteria for notable hedgerows.
	Habitat structure; for example, rivers (Derbyshire Biodiversity, undated; Cambridgeshire County, 2006)
	Breeding bird assemblages using threshold index scores derived from the NCC 1989 guidelines (Cambridgeshire County, 2006; Dorset Wildlife Trust, 2001; Hertfordshire Habitat Survey, undated). (The NCC bird indices are scored in accordance with abundance in the UK, i.e. rarer species gain a higher index value.) Localities with over 50 breeding species, 60 wintering species or 100 passage species (Cambridgeshire County, 2006)
	Amphibian assemblages >3 native species. (Cambridgeshire County, 2006; Essex Wildlife Trust, 2007)
	Reptile assemblages >3 native species. (Cambridgeshire County, 2006; Essex Wildlife Trust, 2007).
	Insect rich habitats supporting >4 bat species (Hertfordshire Habitat Survey, undated)
	River habitat assessment scores under RIVPAC and Pond habitat assessment scores under PSYM (see http://www.brookes.ac.uk/pondaction/PSYM2.htm!) (Dorset Wildlife Trust, 2001)
	Community Conservation Index (CCI) for freshwater invertebrate communities (Chadd and Extence, 2004) and other indices for ecosystem water quality (British Standards Institute, 2000a&b).
	Invertebrate indices (JNCC Invertebrate Site Register, now on the National Biodiversity Network) (Derbyshire Biodiversity, undated)
Naturalness	Ancient Woodlands (Ancient Woodland Inventories, now available on http://www.magic.gov.uk/)
	Defined NVC communities
	Presence of normally between 5 and 10 plant species, depending on habitat, from a list of indicator plant species to define habitats; for example, woodlands, grasslands, ponds, of county value (Derbyshire Biodiversity, undated; Cambridgeshire County, 2006)
	Index thresholds for indicator plant species present where Local Red Data species may be indexed at double the value (Derbyshire Biodiversity, undated)
	Habitat structure; for example, rivers (Derbyshire Biodiversity, undated; Cambridgeshire County, 2006)
Rarity	County rarity being a species occurring in <1% of tetrads, Scare species <4% of tetrads (Hampshire)
	Presence of rare breeding birds being those with less than 300 pairs nationally (Hertfordshire Habitat Survey, undated) or fewer than 20 pairs at a county level (Derbyshire). Species rarity indices based on presence in percentage proportion of county tetrads; for example, for birds (Hertfordshire Habitat Survey, undated)
	IUCN Red Data Book species (most counties)
	Red Data and Nationally Scarce insects (occurring in <100, 10 km^2) to provide a Species Quality Index (Essex Wildlife Trust, 2007)
	Selected Red listed (JNCC/British Trust for Ornithology, 2002) birds (Hertfordshire Habitat Survey, undated; Derbyshire Biodiversity, undated)

Other qualifying criteria for the identification of sites of county wildlife importance draw from those of the Joint Nature Conservation Committee (JNCC) in qualifying the primary Ratcliffe criteria in an empirical manner. Some examples of these criteria are given in Table 3 with reference to some of the counties that cite such criteria in a set of comprehensive guidelines for site selection.

The criteria listed above provide a few examples of the approaches taken at the local scale for ecological evaluation. The appropriate site selection guidelines or criteria for the respective county, borough or city should be consulted for the complete set of criteria used to identify sites of wildlife importance. As such, these approaches offer techniques appropriate for the region to enable a more objective assessment of ecological resources at the local scale.

However, as with using SSSI site-selection criteria as a guide for national importance, some site selection guidelines for local sites should be subject to informed interpretation. For example, the Devon Local Sites Manual gives guidance for reptile sites based solely on the diversity and rarity of the species present and not the population size. This could therefore exclude sites with large populations of individual species, one of the criteria for the identification of key reptile sites (Froglife, 1999).

A number of additional issues arise in relation to the above criteria and the scale at which they are relevant.

•	Ancient woodlands under 2 hectares are generally not included in the UK Ancient Woodland Inventory. Where woodlands may be affected by proposed developments, there is no substitute for a field assessment based on the floristics of the woodland surveyed during the appropriate season. A high value should be placed on most ancient woodland remnants for their inherent diversity, ecological potential and their cultural and historical significance.
•	The NVC, in adopting a level of community division appropriate and manageable at the national scale, does not describe all plant communities of ecological interest (Rodwell et al., 2000). Communities that may not be well described include some transitional types, some locally distinctive communities (e.g. hornbeam [Carpinus betula]-dominated woodlands, meadow barley [Hordeum secalinum] grasslands in south-east England) or rare and unusual communities either not defined or not detected at the national survey level (e.g. lime [Tillia spp.] woodlands, brown sedge [Carex disticha] swamps). In such cases, experience, judgement and local knowledge needs to be brought to bear in the assessment process. Reference to other detailed community classification systems based on floristics can be helpful; for example, Peterken's (1993) woodland stand-types, or river systems according to Holmes et al. (1999).
•	The use of precise numerical criteria, or indices for the evaluation of size, diversity or rarity, raises the issue of the cutoff levels that arise whereby sites apparently failing to reach the target threshold level may be rejected for further consideration, as noted by the Essex Wildlife Trust in their Site Selection Criteria. While numerical criteria can provide a guideline reference scale, professional judgement should be applied with attention paid to as many species or species groups as appropriate; that is, a site evaluated as of low interest with respect to higher plants may be of particular importance for birds or other groups such as invertebrates, lichens or fungi.
•	Numerical estimates of population based on tetrad counts or sampling estimates of population size (e.g. for bird numbers in the UK) should take account of the date of the surveys and the trends in population change since the survey data were collected. The importance of trend data, being for example one of the key criteria for the listing of bird species in the Red and Amber categories for UK birds of conservation concern (JNCC/British Trust for Ornithology, 2002), or the listing of habitats and species in BAPs, is considered further below.
•	The evaluation of habitat mosaics and networks might also be considered to rely heavily on professional judgement, although it is often possible to evaluate the component habitats based on the criteria in Table 3. In addition, some methods aimed at qualifying habitat criteria for certain species—for example, great crested newts (Oldham et al., 2000) and bats (Wray et al., 2008)—also provide some measure of general habitat quality and biodiversity in the relevant locality, from the way the index of habitat suitability is calculated by scoring for the presence of suitable landscape and habitat elements, particularly those providing connectivity.

Ecological evaluations based on species rarity require a good estimate of population at the local level. The quality of information varies from county to county but there are a number of recent publications giving good data on particular groups for certain counties (for example, Killick et al., 1998; Morris, 1998; Murray et al., 1998; Brown, 2000; Green et al., 2000; Green, 2002; Hewlett, 2002; Stott et al., 2002; Wood, 2007). For birds, the local bird club may hold suitable unpublished local data.

BAP Habitats and Species in Ecological Evaluation

The status of the population as measured from static population estimates giving estimates of abundance and rarity is just one factor, and particular attention should be paid to the nature of trends over time in the abundance and distribution of habitats or species populations.

Population trend data, one of the key criteria for Red Data listings (Nature Conservancy Council & Royal Society for the Protection of Birds, 1990), the Red and Amber bird lists (Joint Nature Conservation Committee/British Trust for Ornithology, 2002) and BAP priority habitats and species, provide an additional important element for assessment. However, as noted in the IEEM (2006) guidelines, the presence of some local BAP species and particularly habitats needs to be judged in the appropriate context. Secondary woodlands, for example, can be of prime importance in urban areas but may be of less significance in the rural landscape while urban gardens are clearly entirely reliant on context.

The selection of national and local BAP habitats and species relies, in part, on factors such as population status and distribution, but emphasizes the rate of change (see the criteria in the latest species and habitats review by the Biodiversity Reporting and Information Group, 2007). The BAP list therefore constitutes the best statement available on vulnerability, fragility and, in some cases, incipient rarity, of those species or habitats outside scheduled sites that are currently in need of conservation action in the UK.

With the legal requirement in the UK of all public authorities to have regard for biodiversity, ecological evaluations and impact assessments (ecological impact assessment) must therefore take account of the presence of any and all BAP species and habitats that may be affected by the development proposal so that planners can be informed of the implications for biodiversity as required by PPS 9 (para. 1.vi as quoted above) and whether any predicted adverse impacts may be adequately mitigated. This requirement has recently been tested by judicial review in a case concerning development affecting a characteristic invertebrate community of brownfield land in the Thames Estuary.4

The Ecosystem Approach

Further advice in PPS 9 refers to the ecosystem approach, whereby development needs to address the issue of sustainability and the conservation of resilient and functioning ecosystems. The importance of ecological networks is also emphasized. This offers, in addition to the geographical reference scale for assessment, the unit of the ecosystem for the appraisal of potential effects upon biodiversity from any given development (DEFRA, 2007b). A broader approach to assessment that includes a wider reference to ecological networks and ecosystem services is considered below in the Additional Criteria section.

Three examples arising from the recent experience of the author in impact assessments (from ongoing projects not yet fully in the public domain) for development with respect to the functional ecosystem approach are given in Boxes 1 –3 to illustrate an approach to the assessment of biodiversity at the local scale and the issues that arise.

Box 1. Riverside Infrastructure Development

Development proposals were made affecting a small section of river and river bank in the tidal reach of one of south-west England's ria rivers. Of the ecological receptors considered, two groups came under particular focus in the assessment of local biodiversity, bats and wading birds.

The ecological appraisal indicated that a small number of horseshoe bats (Annex II EC Directive) frequented this section of river and its riverside buildings but at numbers below the level of county importance. However, previous radio-tracking studies indicated that this section of river offered only one of two crossing points for bats commuting from a major roost, notified as a European Directive SAC. This key attribute of the site in a regional network for bats was considered to increase the value of the site to that of county importance (Garland & Markham 2008 consider the legal background to the protection of such commuting corridors).

Similarly, a small population of wading birds were under potential impact from the development, using this section of the river shore mainly as a high-tide roost. Numbers of wintering redshank varied between 20 and 30 birds. This small population approximates to 1% of the county population. Redshank is included in the British Red data bird list in view of the importance of the UK for the migratory and over-wintering population. A second factor of relevance was that, given the geomorphology of the river system, the peak redshank count in the proposed development area represented almost 50% of the wintering population for the river. The presence of up to 4 greenshank was also of significance, given the low national, threshold index of 6–9 birds to register site importance for this species.

Available research based both on modelling and field studies (Sutherland, 1996; Burton et al., 2006) indicate that birds displaced as a result of habitat loss are most likely to suffer reduced fitness, increased mortality with a consequent reduction in regional populations due to competition for resources. This reinforces the commonly adopted precautionary principle that adjacent suitable habitats are similarly at carrying capacity and that displaced species cannot simply be absorbed elsewhere.

At the river ecosystem level, therefore, the development could, without adequate mitigation, result in a highly significant adverse impacts with implications for the local and possibly regional populations of bats and wading birds.

This was accordingly registered in the EIA and strategies for mitigation considered.

Box 2. Proposed Transit Development

A transit development has been proposed for an abandoned railway line. The environmental assessment for the scheme recorded some small patch localities of basic grassland with a few characteristic indicator species present such as upright brome (Bomus erectus), kidney vetch (Anthyllis vulneraria) and yellow wort (Blackstonia perfoliata). However, the vegetation community was, overall, dominated by coarse grasslands, tall herb and scrub comprising common to abundant plant species and the number of indicator species of diverse basic grasslands in any one area generally fell below the commonly adopted numerical criterion for the identification of grasslands of county value (5–6 indicator spp.).

However, a more detailed survey of the whole alignment, treating this as a linear ecosystem, recorded a total of 12 characteristic indicator species of basic grasslands, and the results suggested that ecological succession and the closure of scrub had punctuated a formerly diverse continuum of basic grasslands on the railway verges.

Given this ecological potential and the additional value of the former railway as a wildlife corridor and habitat mosaic, in particular its scrub formations, the site was assessed as of county value and a comprehensive mitigation package put into place involving retention and management of verge habitats where possible on the site, the creation of compensation basic grasslands off-site, equivalent in area to the total area of basic grasslands lost to the development, and a long-term management programme of enhancements to adjacent basic-grassland County Wildlife Sites, thereby enhancing ecological core areas and habitat networks at the local landscape scale.

Use of the corridor for animal movements (e.g. by badgers and foraging bats) also reinforced its valuation as a key linear system and mitigation was directed into preserving, as far as possible, this function.

Box 3. Greenfield Housing Development

A major housing development was proposed for a greenfield site in southern England, one of many such developments proposed in areas of arable and mixed farmland. While the site overall, being dominated by intensive arable farming, was of low habitat diversity, the bird community of arable farmland was typical with breeding species such as rook, barn owl, cuckoo, skylark, yellowhammer, reed bunting, tree sparrow, linnet and turtle dove with occasional wintering flocks of lapwing and golden plover, depending on the cropping regime of the arable land. Brown hare was also present, another typical species of the farmland landscape.

An ecosystem approach to evaluation in this context is problematic given the regional continuum of the agricultural landscape, although sub-elements such as the hedgerows, small watercourses and old field ponds can be addressed in terms of tree and shrub diversity and habitat structure, amphibian populations and the connective function of the hedgerows and field verges as part of the larger landscape ecological network. Master planning addressed the need to retain these features in a connected matrix of extensive public open spaces that linked to the wider countryside beyond; in particular, targeted at a great crested newt metapopulation. However, such on-site mitigation would not address the likely losses of the farmland bird and mammal community.

Bird surveys recorded a number of BAP species present as breeding populations. For some (e.g. song thrush, dunnock)—the development with the proposed mitigation strategy of extensive areas of open space being retained, encapsulating the key wildlife corridors—no significant impacts were anticipated. For other species, however (e.g. grey partridge, yellowhammer, turtle dove), it was concluded that the land-use change would probably result in the loss of the species from the site.

Developments in agricultural landscapes thus raise the issue of how to evaluate, for example, the significance of presence of one or possibly a number of BAP species dependent on farmland in the context of their highly probable loss from the area following development. At what level is it reasonable to require mitigation for the loss of low numbers of BAP species bearing in mind that a financial cost to ecological benefit ratio is relevant? An approach to this is considered below.

Assessments of Farmland Mosaics using Birds as Indicators

Birds are commonly used as surrogates for biodiversity, and the decline of the bird assemblage of farmland is of concern throughout much of Europe (Gregory et al., 2005). In the UK, the NCC (1989) diversity indices for the calculation of breeding bird community index thresholds are based on 1984 population estimates and their SSSI Guidelines do not give a threshold index for farmland bird communities as such (Lowland Damp Grasslands and Scrub with threshold indices of 16 and 15, respectively, contain a few species in common; see Table 4). Given the notable declines in the UK populations of farmland birds, the 1989 indices require re-calculation after which a threshold value may be set for community diversity of farmland birds.5

Table 4. Proposed indices for the assessment of farmland birds

Species	Status	Approximate UK population estimate	Updated NCC index	Population on site (breeding pairs)	Combined NCC/site-population index
Turtle Dove^a	BAP	44,000	3	2	6
Cuckoo^a	BAP	10,000–20,000	3	2	6
Barn owl	Schedule 1	4,000	4	1	4
Kestrel	BAP	36,800	3	1	3
Skylark	BAP	801,000–1,785,000	1.5	21	31.5
Linnet	BAP	556,000	2	4	8
Yellowhammer	BAP	792,000–1,200,000	1.5	18	27
Tree sparrow	BAP	68,000	3	1	3
Reed bunting^a	BAP	200,000	2	1	2
Combined Community Index					90.5
Note: ^aNCC community indexed.

In the example shown in Table 4 (with farmland bird species taken from the development site considered in Box 3), the original NCC indices are modified so that all BAP species may be indexed by commencing the abundance scale for those species still with over 1 million breeding pairs nationally with an index of 1. A total site index of 26 for the breeding bird community is scored, taking only those species considered at high risk of loss to the development, which, even allowing for the increase by one of each abundance category, compares well with the threshold indices for lowland damp grassland and scrub habitats (NCC, 1989). Thus a reasonable case for mitigation is established.

Community diversity is but one measure of ecological value; population size is also relevant. Where significant populations of BAP species are present, even in sites of low habitat diversity such as in areas of intensive farming, these should be considered as being of importance for local biodiversity or relevant at regional scales; for example, a site with over 4 pairs of breeding turtle dove or over 12 pairs of breeding yellowhammers. Where it is considered that such populations would be lost to a development, it would be necessary and in keeping with the aims of the UK BAP and PPS 9, for such losses to be mitigated. A combined index to take into account both community diversity and population size of BAP species present might assist in the evaluation, and in determining the need for mitigation. Here the species index is adjusted by a multiple representing the size of the population on the development site, as shown in the final column of Table 4. Such an index would encompass assessments of those areas of relatively low habitat diversity but where a few specialist BAP species may be present and reliant on resources in the proposed development site.

A threshold value of around 25 for the combined diversity/population size index could be considered for mitigation action to compensate for losses to a BAP species population or species in a community of relatively low-species diversity. Thresholds are clearly a matter of judgement and, as such, should be flexible according to the status of local populations and the contribution those populations make to local or regional biodiversity. Thus, different thresholds may be adopted for local (county) scales depending on the abundance or rarity of the species (where known) at that scale. Consultation with local specialists and conservation agencies will inform this judgement and assist in developing a consensus. As noted above, evolving case law may also inform on the need for effective mitigation where BAP species are adversely affected by development.

Where mitigation is considered necessary, it should be noted that, as considered by Latimer and Hill (2007), certain impacts cannot readily be mitigated by on-site, small-scale habitat retention or manipulation (issues relating to habitat substitution are considered further below). Losses of estuarine habitat (Box 1) and the conversion of farmland to urban and sub-urban land use (Box 3) are two examples where a more strategic approach to offsite compensation is required. For the farmland birds displaced by construction this would require habitat enhancements of farmland elsewhere in the region, using techniques adopted under Environmental Stewardship (Winspear & Davies, 2005), preferably in advance of any loss of habitat to development so as to avoid temporal losses to populations. Effective mitigation for such losses to farmland species from urban developments is necessary to support the UK national DEFRA Public Service Agreement targets for the conservation of 20 species of farmland birds, with current programmes mainly dependent on central government funding.

Additional Criteria

Capacity for Substitution

Earlier ecological assessments in the UK under New Approach to Appraisal (DETR, 1998) required some consideration of the capacity to re-create habitats lost to transport schemes, aiming to take account of the significance of the loss of an irreplaceable critical natural capital and to assess residual impact assuming mitigation to be in place. Habitats deemed irreplaceable under realistic timescale (e.g. ancient woodlands, peatlands) are generally given a high evaluation score, while conversely re-creatable habitats such as secondary woodland or coarse unmanaged grasslands attract lower scores in the evaluation matrix. However, it is necessary to consider the importance of those habitats considered re-creatable in the appropriate context (e.g. in urban environments, these may require a higher ranking), and to take account of the timescales required for such habitats to mature and the implications for any dependent species in the interim period.

Furthermore, while it may be possible to reinstate the core structure or dominant species of a particular habitat considered to be re-creatable in new compensation areas, a probability of success and a qualification of timescales required to produce the compensation habitat needs to feature in the impact assessment. Clear and precise survey data regarding the biodiversity inherent in any habitat lost are needed to inform this assessment in addition to the relatively extensive literature on the varying degrees of success or failure of habitat creation projects (for example, Mitsch & Wilson, 1996; Gilbert & Anderson, 1998; Box, 2003), so that adequate compensation can be provided. Reedbeds comprising Phragmites australis are an example of a habitat that can often be successfully re-created, increasingly at landscape scales with benefits for a number of ‘flagship species’. However, examination of a natural reedbed shows the presence of a ‘ground flora’ of submerged and low, emergent marsh plants, a component usually absent in re-created mono-specific stands. Similarly some scrub-grassland mosaics may have developed on old, species-rich grasslands and careful survey is needed to detect the complete range of species present. While scrub and coarse grasslands are relatively easy to establish, they may lack the biodiversity present in older natural mosaics.

It is perhaps telling that capacity for substitution is no longer part of the TAG appraisal process. Careful survey of habitats subject to impact, and a transparent risk assessment attached to the likely success of proposed recreated habitats, are essential for any assessment of residual impact. While the former should be readily achievable, the latter is fraught with uncertainty, a factor militating for mitigation to be undertaken well in advance of any loss of ecological resources due to development.

A perceived capacity for substitution therefore has no place in the valuation of habitats but, following survey and evaluation, can inform the impact assessment and appropriateness of mitigation.

Ecological Networks

Conservation planning has addressed biodiversity at varying scales; from a coarse-filter approach where biodiversity objectives are deemed to have been captured by the notification of large reserves or whole ecosystems at regional landscape scales, to a fine filter that identifies the requirements for specific species, usually at more local scales. For some ‘umbrella’ or ‘keystone’ species where significant habitat areas and supporting inter-related communities are required, the conservation outcome may be similar (for a review of these concepts, see Vold & Buffett, 2008). Intermediate scales are recognized (Hunter, 2005) for habitat patches such as ponds or standing dead timber and, essentially, multi-scale assessments are required (Poiani et al., 2000; Noon et al., 2003) for conservation planning.

These approaches, however, are considered to have failed to adequately address the issue of wildlife corridors and spatial ecological networks in human-dominated landscapes (Simberloff, 1998; Cowling et al.; 2004). Such a requirement has long been recognized and the debate, which originally centred on the efficiency of linear landscape connections in relation to species' dispersal capacity (for example, Wood & Pullin, 2002), has moved on to a consideration of ecological networks where a matrix of interconnected habitats comprising core areas, patches and corridors of variable width are considered to provide the best option for species to move across landscapes where human activities predominate (Poiani et al., 2000; Bennet, 2004; Opdam et al., 2006).

Biodiversity assessments now need to consider the need for and role of ecological networks in relation to the dispersal behaviour of target species in the area under scrutiny. Networks are required to function at all landscape scales, a need more pressing in view of climate change. Retention of habitat patches and the integrity of wildlife corridors at local scales will contribute to the efficiency of the wider network. Thus, biodiversity assessments for EIA need to identify any elements of an ecological network that may be present, both for its intrinsic ecological qualities at that scale using the criteria considered above, and the contribution it may make to regional connectivity. Considering the development examples given in the boxes above, Box 1 cites an assessment of the site in the context of the regional dispersal network for bats, and Box 3 cites an assessment of hedgerow networks and habitat patches with reference to a great crested newt metapopulation that extended from within to beyond the development site.

Ecosystem Services

The assessment of ecosystem services, particularly in the application of economic evaluations, is an emerging area of ecological assessment, although the details of monetary evaluations may be largely beyond the immediate expertise of many ecologists. However, qualitative assessments based on an ecological knowledge of ecosystem qualities and processes can be framed by ecologists in the assessment and evaluation process, and can contribute to more detailed economic analyses where needed.

There are a number of correlates between biodiversity and the capacity for ecosystems to provide the services required, for example, for attenuation and storage of rainfall, pollution control and bioremediation, climate regulation and the sequestration and storage of carbon (e.g. by woodlands or peatlands). Intensive land use and the simplified ecosystems that can arise, for example, from overstocked and overgrazed hill and lowland grasslands, have deleterious consequences for the efficiency of ecosystem regulation; for example, in relation to water infiltration, attenuation, nutrient cycling and soil erosion (Ball et al., 1997). In general, the more diverse the vegetation structure, the better the control over surface run-off and soil erosion (Quinton et al., 1997). Hedges and field margins can be valued for their wildlife and landscape characteristics, as wildlife transmission corridors in an ecological network but also for their role in the control of wind and water erosion, local climate control and as habitat for agents of plant pollination or biological control for agricultural pests (Forman & Baudry, 1984). Diversity can enhance primary production (Vitousek & Hooper, 1993) and the decomposer cycle (Swift & Anderson, 1993). The role of natural wetlands in water attenuation, storage and cleansing, the contribution of rivers to estuarine and marine productivity, and the function of biogenic reefs and coastal marsh in sea defence are similarly well-documented ecosystem services with clear economic benefits as well as distinctive biodiversity attributes.

Context is also relevant to the assessment of ecosystem services where the positive effects of, for example, woodlands, wetlands and watercourses for local climate regulation and regulation of surface run-off, as well as their aesthetic qualities, would be valued more highly in urban environments.

However, the concept of ecosystem services should not be regarded as a complete surrogate for biodiversity since there does not appear to be a complete or linear relationship between the efficiency of some examined processes and some of the more biodiverse systems (Schwartz et al., 2000). However, what may appear to be redundant elements or species in the system may confer resilience (Cairns & Neiderlehner, 1993) and these elements may also confer value in terms of serependipitous, unrealized assets or intrinsic aesthetic qualities.

In general, it is the coarse or regional scale at which ecosystem services are best considered (Swift et al., 2004). However, local assessments apply in order to determine effects on local components of larger ecosystems and how this may affect the whole system; for example, developments along watercourses, flood plains and in river catchments, or the implications of effects on small habitat patches or corridors on the functioning of ecological networks and the resilience of meta-populations.

A consideration of ecosystem services in the areas of regulation and support of natural cycles and climate, provisioning of materials or foods, aesthetics and cultural requirements (DEFRA, 2007a&b) is therefore highly pertinent to the assessment of natural and biodiverse systems in EIA and the assessment of ecological risk (Cairns & Neiderlehner, 1993), and for land-use planning (Haines-Young & Potschin, 2007).

Conclusions

It is the role of EIA practitioners to advise planners as to the environmental implications of development. Specialists in ecological impact assessment should provide sufficient information on the implications for biodiversity to inform the decision-making process under the requirements of government policy in PPS 9 with reference to the likelihood of ‘significant harm to biodiversity’, whether such harm can be ‘adequately mitigated against’ and how ‘appropriate compensation measures should be sought’. With this information, planners can then be better able to consider whether ‘that significant harm cannot be prevented, adequately mitigated against, or compensated for, then planning permission should be refused’.

The matrix approach (as in TAG) to ecological impact assessment has some value in providing a consistent national framework for ecological impact assessment but lacks sensitivity, focuses attention on notified sites and species for which there is often a good baseline of information, and, conversely, tends to result in the under-valuation of ecological resources at local scales. A greater degree of refinement is needed for the assessment of ecological resources at these levels.

Criteria for ecological evaluation may be drawn with reference to those adopted for the selection of sites of national and international ecological importance and increasingly being applied at county levels by The Wildlife Trusts or local biodiversity partnerships for the selection of county wildlife sites. Reference should be made to these local criteria for ecological evaluation and impact assessment. These will range from local applications of the Ratcliffe criteria to the application of indices of diversity or rarity tuned to local conditions where factors such as local population estimates or community distributions are known.

While much can be gained from consultation and data searches from local or national biological records to provide a reference baseline against which assessments can be made, there is no substitute for site-specific data derived from field surveys undertaken during the appropriate season to facilitate classification at community level, evaluation and assessment of potential impacts While the use of surrogate indicators such as higher plants, birds and mammals can provide representative indicators for biodiversity, and their ecological requirements often reflect a landscape scale of habitat pattern, in most cases it is important that other groups should be assessed to encompass a more complete spectrum of biodiversity at varying geographical scales from micohabitats to regional networks. The importance of local biodiversity reference data for site evaluation re-enforces the need to provide data collected in ecological impact assessment to the appropriate record centre.

The geographical reference scale against which the site evaluation is performed should always be stated; whether the criteria refer to pertain to the national, county, natural area, or ecosystem level. Where it is possible to identify finite ecosystems (e.g. a woodland, river or marsh), potential impacts on biodiversity can then be evaluated with reference to ecosystem integrity.

In addition to static measures of abundance, trend data provide one of the key criteria for assessment, both as a possible predictor of incipient scarcity or rarity and as a guide to conservation imperatives. Thus the presence of BAP habitats and species on sites undergoing assessment needs to be taken into account, evaluated against a stated geographical baseline and, where significant adverse impacts are anticipated, proposals made for appropriate mitigation to comply with the objectives of the Plan. Other trend data may be available for a similar analysis; for example, the Red and Amber lists for birds of conservation concern. The use of abundance indices as used in the assessment of bird communities for the selection of SSSI can provide a tool to take into account current population levels and population trends, although the setting of thresholds for mitigation action is a matter of judgement.

Where a final assessment of residual impact has taken into account a proposal for the re-creation of habitats lost to development, such judgements must be made on sound data derived from a thorough field survey and assessment together with a stated assessment of probability that the re-created habitat will succeed in its intended function. This will often be a matter of professional judgement in the context of the precautionary principle. In general, habitat creation, including translocations, should be regarded as a last resort, and anticipated outcomes should not influence the original assessment of value.

Biodiversity assessments in EIA need to identify the presence of spatial ecological networks or elements of these and appraise their intrinsic importance according to the criteria considered in this paper, and the contribution that the local elements make to the regional matrix of core areas, habitat patches and wildlife corridors. The anticipated impacts of any loss to those local elements should be assessed in terms of the functional integrity of the network, both locally and at appropriate wider scales.

The concept of ecosystem services, while not by any means providing a novel appreciation of the role that natural ecosystems have in providing life support systems for human populations, heralds the application of environmental economics to biodiversity assessments. Given some correlation between biodiversity and the capacity or efficiency of ecosystems for provisioning, for regulatory and support systems, and their amenity and cultural significance, an appraisal of ecosystem services can add value to biodiversity assessments for land-use planning.

Acknowledgements

The author is grateful for the support and assistance from Richard Graves, Ecology Director, Faber Maunsell, Peter Burston, Senior Ecologist, Faber Maunsell, Dr Lincoln Garland, Biodiversity by Design and the stimulating comments of two anonymous referees.

Notes

1. The appraisal approach has gained national acceptance for a range of developments and is now published as Transport Analysis Guidance (TAG) (http://www.webtag.org.uk/webdocuments/doc_index.htm, accessed June 2009). TAG Unit 3.3.10 deals specifically with ‘The Biodiversity Sub-Objective’.

2. This is available on the JNCC web site with updates (http://www.jncc.gov.uk/page-2303#download, accessed June 2009).

3. Reference can be made to the complete JNCC SSSI guidelines for further examples of selection criteria for particular habitats or species groups in the UK, and to the IUCN web site (http://intranet.iucn.org/webfiles/doc/SSC/RedList/redlistcatsenglish.pdf) and the Convention on Biological Diversity web site (http://www.cbd.int/indicators/) for criteria relating to biodiversity assessments at the national and international scales.

4. See http://www.lawreports.co.uk/WLRD/2008/QBD/feb0.3.htm (accessed June 2009).

5. Current estimates for UK populations of these species are available on the British Trust for Ornithology web site (http://www.bto.org/birdtrends2005/discussion43.htm, accessed June 2009).