Critical habitat under PS6: the five-criterion test step by step

Qualifying as critical habitat is the most demanding shift in PS6. On one side, natural habitat where no net loss is the aim. On the other, critical habitat where net gain becomes mandatory, where the very existence of the project must be justified, and where a poorly supported file blocks financial close. The problem is that this qualification rests on five criteria and precise thresholds that many consultants misapply. This article unpacks the five-criterion test step by step, gives the thresholds where they exist, and details the errors that cost a study revision.

Critical habitat: a category, not a nuance

PS6 classifies all habitat into three families: modified, natural, critical. Critical habitat is not a fourth type. It is a subset of modified or natural habitat that exhibits high biodiversity value. The same site can therefore be both natural and critical, or even modified and critical.

The standard states: "Critical habitats are areas with high biodiversity value" (IFC PS6, paragraph 16). It then lists five families of values. The presence of one alone is sufficient to qualify the habitat as critical. There is no hierarchy among the criteria: none is more decisive than another.

First practical consequence. Qualification is binary. A site is in critical habitat or it is not. There is no semi-critical habitat that one might negotiate at the margin. All the subtlety of the work therefore consists in passing each biodiversity value through the sieve of the five criteria, without forgetting one, and documenting the result.

Second, more counter-intuitive consequence. Qualification is independent of the project. It does not depend on the footprint, nor on impacts, nor on the mitigation strategy. A site is in critical habitat because it harbours certain values, whether the project is a power station, a road or a mine. GN6 is explicit: "Clients should not claim they are not in critical habitat based on the project footprint or its impacts" (IFC GN6, paragraph GN60). This is error number one in weak files: concluding the absence of critical habitat because the project "touches little".

Fix the right area of analysis before any test

Before applying a single criterion, one must decide where to look. This is the step that hurried consultants skip, and it is the one that determines all the rest.

GN6 requires defining an ecologically relevant area of analysis for each species and each ecosystem concerned. This area is not the project footprint. It follows the distribution of the species and the processes that maintain it. It may follow a watershed, a large watercourse, a geological formation. It often extends beyond the project's area of influence. The guidance note states bluntly: critical habitat assessment should not be limited to the project site.

The issue is methodological, not cosmetic. Take a fish species with restricted distribution in a basin. If the area of analysis stops at a few hundred metres from the water intake, the species appears anecdotal. If it covers the basin, one discovers that the site concentrates a significant portion of the global population. The same site, two perimeters, two opposite conclusions. This reasoning joins that of environmental flow, where the life of a bypassed reach depends on a well-defined perimeter.

Best practice unfolds in three stages that GN6 structures: a documentary review and expert consultation, field data collection, and then the actual determination. The first two steps are conducted without asking whether the project will have an impact. One first seeks to understand the landscape, not to defend a conclusion. This scoping extends the biodiversity issues that every project owner must anticipate from the study phase.

The five criteria, unpacked one by one

GN6 specifies that numerical thresholds have been defined for the first four criteria, aligned with the IUCN Key Biodiversity Areas standard. The fifth criterion has no numerical threshold. Here is the test, criterion by criterion.

Criterion 1: Critically Endangered or Endangered species

This criterion targets species classified as CR (Critically Endangered) or EN (Endangered) on the IUCN Red List. The mere presence of the species is not sufficient. A significant concentration is required.

The GN6 threshold combines two cumulative conditions: an area that harbours at least 0.5% of the global population of the species AND at least 5 reproductive units. A reproductive unit is not an isolated individual: it is the minimum number of mature individuals necessary for a reproductive event, for example five pairs. The criterion can also apply to a Vulnerable species whose loss would shift the status to EN or CR. National or regional red lists are taken into account according to the modalities set out in the guidance note.

Two points of vigilance. First, the threshold is cumulative: both conditions must be met. Second, great apes (gorillas, orangutans, chimpanzees, bonobos) are subject to special treatment. Any site likely to harbour them is very probably treated as critical habitat, and an IUCN specialist group must be consulted very early.

Criterion 2: endemic or restricted-range species

Here, GN6 first defines "restricted range" by a limited extent of occurrence (EOO). Thresholds differ by environment.

• Terrestrial vertebrates and plants: EOO less than 50,000 km².
• Marine systems: EOO less than 100,000 km², provisionally.
• Coastal, riverine or aquatic species in habitats not exceeding 200 km in width (a river, for example): global area less than or equal to 500 km of linear extent.

Once the species is recognised as restricted-range, the qualification threshold applies: an area that regularly harbours at least 10% of the global population AND at least 10 reproductive units. Again, cumulative condition. Many files confuse national endemism with restricted range in the PS6 sense. An endemic species of a large country may have an EOO well above 50,000 km² and not fall under criterion 2.

Criterion 3: migratory or congregatory species

This criterion targets concentrations of migratory species (which move cyclically from one area to another) or congregatory species (which gather in large groups predictably). Breeding colonies, migration bottleneck sites, source populations: the guidance note gives several examples.

Two thresholds are provided. An area that hosts, on a cyclical or regular basis, at least 1% of the global population of the species at any point in its life cycle. Or an area that predictably supports at least 10% of the global population during periods of environmental stress. The key point here is temporal. A site may show nothing nine months out of twelve then concentrate a major portion of a population in migration. Hence the importance of inventories covering the right seasonal windows, not a single campaign.

Criterion 4: highly threatened or unique ecosystems

This criterion does not focus on a species but on the ecosystem itself. GN6 refers to the development of the IUCN Red List of Ecosystems, on a logic similar to that for species. Ecosystems that have lost most of their extent or whose global distribution is very limited are targeted. The quantitative reasoning exists, but it is less standardised than for criteria 1 to 3. The demonstration relies on recognised classifications of threatened ecosystems and expert opinion.

Criterion 5: key evolutionary processes

The fifth criterion is the most qualitative. It targets areas associated with key evolutionary processes: physical landscape structures linked to particular biodiversity characteristics, or areas harbouring populations with distinct genetic characteristics. It is the only criterion without a numerical threshold. It is demonstrated by analysis and expertise, not by a percentage. It is also most often forgotten in files, because it does not fit into any numerical box.

Tier 1, tier 2: reading the intensity of qualification

Practitioners often speak of "tier 1" and "tier 2" to describe the intensity of a critical habitat qualification. This vocabulary comes from a two-level reading of sensitivity: the more a site concentrates a high proportion of a population or a rare ecosystem, the more irreplaceable the value.

One must be precise about what the current text says. The June 2019 version of GN6 does not organise the criteria into formal tiers. It sets unique numerical thresholds, aligned with the IUCN Key Biodiversity Areas standard. The "tier 1 / tier 2" shortcut remains a useful reading tool for prioritising the sensitivity of a portfolio, but it does not replace the test of thresholds published in the guidance note. A solid file relies on the exact thresholds (0.5%, 5 reproductive units, 10%, 1%), not on a tier label.

This distinction has operational value. It helps to sort early, in the feasibility phase, between sites where the risk of critical habitat is probable and those where it is marginal. But at the time of formal demonstration, only the documented crossing of thresholds matters, criterion by criterion.

What qualification triggers

Concluding critical habitat is not an observation, it is the opening of a much stricter regime of requirements. PS6 sets the entry threshold. The client cannot launch any activity until it demonstrates four cumulative things.

• No other viable alternative in the region on non-critical habitat. This is the avoidance obligation, first step of the mitigation hierarchy.
• No measurable adverse impacts on the biodiversity values for which the qualification was made, nor on the ecological processes that support them.
• No net reduction in the global or national population of a CR or EN species over a reasonable time period. This condition targets only criterion 1.
• A robust, well-designed and long-term biodiversity monitoring and evaluation programme, integrated into the management system.

And above all, the requirement shifts to outcome. In natural habitat, no net loss is the aim. In critical habitat, the standard is higher: "a net gain is required in critical habitats" (IFC PS6, paragraph 10). This net gain relates to the values for which the habitat was designated. It is formalised in a Biodiversity Action Plan and may involve, if necessary, offsets. Offsetting remains the last step of the hierarchy: it is only discussed after exhausting avoidance, reduction and restoration.

Concretely, qualification as critical habitat transforms the conversation with the lender. One no longer talks about reducing impacts; one must prove that no alternative existed, that values persist, and that one delivers a measurable gain. It is a file of another order, to be anticipated well upstream.

At review, DFI E&S teams verify less the conclusion than the robustness of the path leading to it.

Critical habitat is not declared, it is tested. Three reflexes avoid study revision. Fix the area of analysis first, at the scale of the species not the construction site. Unpack the five criteria without skipping any, applying cumulative thresholds to the letter. Clearly separate qualification, independent of the project, from impact assessment, which comes afterwards.

The right question is not "does my project touch much", but "does this site cross one of the five thresholds". A file that answers this question, value by value, withstands due diligence. A file that concludes the absence of critical habitat on grounds of a small footprint undergoes it, and discovers the net gain requirement at the worst moment, that of financial review.