Plenary09: Making the most of birds as environmental indicators

BirdLife International, Wellbrook Court, Girton Road, Cambridge CB3 0NA, UK, e-mail colin.bibby@birdlife.org.uk

Bibby, C.J. 1999. Making the most of birds as environmental indicators. In: Adams, N.J. & Slotow, R.H. (eds) Proc. 22 Int. Ornithol. Congr., Durban. Ostrich 70 (1): 81–88.

Never previously has there been such an opportunity for biologists to inform and advise environmental decision-makers. Ornithologists have a special contribution to make because of the extent and quality of the information available to us. This paper focuses on environmental indicators – the process by which complex information about the environment can be encapsulated in some simply understood measures. This paper reviews the properties of effective indicators and the contribution that ornithology can currently make. Attention is given to globally threatened species, common bird monitoring schemes, summarising trend data from many species and to spatial questions about the distribution of biodiversity.

Biologists have been pointing to loss of species and habitats for a long while and have contributed greatly to the development of a nature conservation ethic and science. We now live at a time of rapidly growing understanding of the linkage between human development and the environment (World Bank 1997a). This offers an opportunity to bring biodiversity conservation into the mainstream of economic and land-use planning.

The environment can no longer be seen as an infinitely large source of free goods and services. We can no longer ignore the fact that the damage people do to their environment in turn causes damage to the prospects for development. In general, poor people often bear the greatest costs of environmental damage. Development aims to improve the quality of life and help people out of the trap of poverty. This both increases their potential for further environmental damage through greater resource consumption but also increases their ability to afford a healthy environment. Finding sustainable development strategies is a central challenge to mankind. Increasingly, a country’s wealth is viewed as a combination of manufactured, natural and human resources. Growing one at the expense of others is not sustainable (World Bank 1997b).

Aside from the Rio Conference on the environment and development (1992) other recent United Nations conferences have dealt with education (1990), population (1994), women (1995) and social development (1995). Building from the broad agreement from these meetings, the Organisation for Economic Co-operation and Development (OECD 1996) has called on the global development partnership to achieve a set of social, economic and environmental goals. The key point is that factors such as poverty reduction, education, health, the role of women and the state of the environment are inseparably linked (World Bank 1995). Development of goals and plans to achieve them requires knowledge and communication of understanding across these traditionally widely disparate sectors. Indicators are one format for the simplification and transfer of such understanding.

The use of indicators in economic policy making and management is well established at global and national levels and some social factors are consistently measured and reported as a basis for policy making. Most environmental decision making is based on poor access even to what limited information is available (World Bank 1995). This is unfortunate at a time when unprecedented effort is going into the search for sustainable development strategies and plans. Policy formulation will not wait so conservation biologists need to be quick to seize the current opportunity. Countries are producing national environmental studies and plans at a considerable rate (INTERAISE 1996). The role of monitoring within the cyclic and adaptive process of biodiversity planning is becoming ever clearer (Miller & Lanou 1995). Attention is being given to the integration of biodiversity with other sectors such as agriculture (see for instance Anon 1997).

The Convention on Biological Diversity, now ratified by some 172 countries, is the centrepiece of global efforts to ensure sustainability of use of biodiversity and the equitable sharing of its benefits. Numerous decisions have been reached within this mechanism and countries have made some important commitments to the development of indicators (SCBD 1996). Amongst these are ‘to identify indicators of biological diversity and develop innovative methods of implementing Article 7’ (which concerns identifying key components of biodiversity and threats to them). Parties have been called upon to co-operate on a pilot project to demonstrate the use of assessment and indicator methodologies. Parties have also agreed to set measurable targets within biodiversity conservation strategies. Discussions have acknowledged that these steps will require innovation, which provides an opportunity for ornithologists to offer help from their substantial experience in co-ordinated monitoring and data gathering.

Indicators are meant to quantify and communicate complex phenomena, in this case, biodiversity trends and patterns, in a simple manner. Their intention is to help decision makers formulate policy and then modify it in response to changes of the indicators. Public access to indicators allows the public to judge the success of decision making. Indicators are not a substitute for the detailed knowledge needed to assess the causes of changes or to formulate strategies or plans in response. They can be regarded as watch-dogs whose barking needs to be heeded. The use of indicators is familiar in the economic field where a small number of statistics, such as the Retail Price Index or the Public Sector Borrowing Requirement, are widely used by the public and decision makers to track the economy and the political performance of its managers. It neither matters that most people would not know exactly how these figures are calculated nor that they do not represent all the detail of the economy. They are nevertheless used to make decisions in government, business or personal financial matters.

In using a broad definition of the term indicator, I am aware of some risk of confusion. I leave questions such as whether birds are good indicators as meaningless without further definition (indicators of what?). I discuss both spatial and temporal phenomena, though it is mainly the latter which are of policy interest and relevance. The ultimate interest of conservation is the sustainability of use of the environment. Survival of biodiversity is part of environmental sustainability and birds are part of biodiversity. I write later of birds indicating the non-sustainability of European agricultural policy. In inferring non-sustainability from bird counts, I make no comment about whether social and economic dimensions may or may not be sustainable – this would obviously require other data. In using birds as a dimension of biodiversity, I make no comment on the extent to which other biodiversity might also have declined as a result of European agricultural policies. This is an interesting question which would also require other data.

Indicators do not make much sense without reference points against which the significance of change can be assessed. One kind of reference is the baseline, which is a starting point at some time or state – such as the area of wetlands now or at the start of the modern age. Thresholds may be formalised to set some stage at which alarm bells ring – for instance that a species has become threatened. The most valuable kind of reference point is the target, set as an agreed measurable end point against a time-table. Biodiversity conservation has only recently come to the discipline of setting targets for its activities. The combination of targets and indicators provide a means for policy makers and civil society to agree objectives and to observe progress towards them. Biodiversity planning can then become an adaptive process, learning as it goes (Miller and Lanou 1995).

It has not proved easy to develop indicators for biodiversity. Like all indicators, they need to have a variety or properties which are often conflicting (see Hammond et al. 1995 for an overview):

• Quantitative – so that they can be formally measured without ambiguity

• Simplifying information – because the complexity of biodiversity and knowledge about it has to be conveyed briefly if it is to have impact

• User driven – because the stakeholders involved will have to commit to plans and targets being monitored and will often collect the relevant data

• Policy relevant – because the point of indicators is to return signals to policy makers about their impact and effectiveness

• Scientifically credible – so that the methods and inferences are sound

• Responsive to changes – so that trends are picked up without undue delay

• Easily understood – so that non expert policy makers or members of the public have a sense of ownership and sympathy

• Realistic to collect – because biodiversity is often poorly known and expensive to study and there is a limit to society’s willingness to pay field biologists

• Susceptible to analysis – so that the data can be disaggregated in pursuit of possible causes of trends

While some of these properties may seem rather obvious when listed, it is not difficult to think of potential indicators which have some appealing properties but turn out to fail seriously on other criteria. Finding indicators which meet most criteria is surprisingly difficult.

Work by OECD (1993) has proposed a widely accepted taxonomy of environmental indicators, dividing them into three classes:

(2) State indicators measure the actual condition of the environment. Examples would include pollutant burden measured in the air or rivers, area of forest, trends in the abundance of species or number of threatened species. Those to do with the ability of ecosystems to provide services, including biodiversity, have proved particularly hard to formulate and develop.

(3) Response indicators show what has been done to alleviate pressures by controlling problems or benefiting the environment. Examples include proportion of sewage treated, expenditure on pollution control, or area protected. In general, they measure progress towards standards or compliance with regulation. They do not say much about what is happening to the environment because the standards or laws may or may not be sufficient. There may be a high percentage of land under protection, but that does not necessarily mean that species and habitats are secure.

Three general features of OECD’s proposals are striking. First, pressure indicators are more easily formulated and measured than either state or response. Secondly, a high proportion of cited examples cover pollution and its abatement which can readily be measured as pressure (i.e. at source), as state (i.e. in the environment) or as response (in costs and facilities for control). Finally, within the environment, biodiversity gets very short shrift and as a result, many dimensions of non-sustainable land-use, such as habitat deterioration and its effects on species can evade capture (e.g. MacGillivray 1994; see also WWF 1998).

Some examples of potential indicators for biodiversity in the PSR framework are summarised in Table 1. Indicators can be formulated at a variety of scales. In the context of the Convention on Biological Diversity, the most pressing level is national. It is at this level that most biodiversity planning will be done and this is the main level at which it is possible to integrate biodiversity into all sectors of the economy. Regional and local plans will require a finer detail to support indicators. Table 1 has been synthesised from a variety of sources (see Hammond et al. 1995; MacGillivray 1994; OECD 1993) but its coverage of biodiversity is fairly typical. It is appropriate then to review the contribution that ornithology might make in such a framework and the problems that need to be overcome.

Threatened species represent one class of urgent conservation priority because the options for keeping them on Earth are fast closing. Official lists are maintained by IUCN–The World Conservation Union (IUCN 1996) and revised every three or four years. Responsibility for birds is delegated, by IUCN, to BirdLife International. Birds are more thoroughly documented than any other Class with all species having been reviewed and those that are classed as threatened supported with documentation (Collar et al. 1994, 1988). For Africa and the Americas, extensive documentation has been published in Red Data Books (Collar et al. 1985, 1992) and work on Asia is well advanced with preliminary publication due in 1999. Mammals were fully reviewed for the first time for the 1996 list but the published record provides minimal documentation, well below the agreed standard recommended for Red listing (IUCN/SSC 1994). No more than partial lists exist for other taxa. Many countries have official Red lists though these do not always make the fundamental distinction between globally threatened species and those which are merely locally rare or threatened in the country concerned.

The list of threatened birds makes shocking reading with nearly 12% of birds at risk of extinction (Collar et al. 1994). This represents a likely rate of extinction several orders of magnitude above the background rate from palaeontological history (Lawton & May 1995). Relatively few threatened species occur in north temperate latitudes where, sadly, the majority of ornithologists and conservationists still live and work. In broad terms we understand the underlying causes. Most devastating has been the introduction of predators and herbivores to islands. As evidence grows, estimates of losses in the Pacific continue to rise but are now known to exceed 2000 species since human colonisation (Steadman 1995). More recently, human transformation of continental habitats has come to the fore with a combination of habitat modification and fragmentation. A high proportion of threatened species have limited ranges in centres of endemism which have lost very high proportions of their native vegetation (ICBP 1992; Stattersfield et al. 1998).

How well do threatened birds match up as effective indicators against the properties suggested above? Undoubtedly, they score quite well on simplification with a very plain message that is easily understood and attracts sympathy. With clear and agreed criteria, they are increasingly scientifically credible though we still carry a legacy of alarmist predictions about extinction rates expected by the end of the century which have not come true. The cautious message is shocking enough without need to exaggerate our case.

Threatened species lists work less well as indicators in some other respects. They are not very responsive to change being likely to take decades to alter as species become extinct or are saved by effective interventions. Most changes so far have come from change in criteria or knowledge rather than the status of species. Overall, they do not fit that well into a policy framework, because they are mainly rare and peculiar and are threatened by a wide variety of factors. As a result, it is not always easy to link policies in different economic sectors with likely responses from globally threatened species. The final weakness with this indicator is the degree to which it is at the moment quantitative beyond the length of lists. Although birds are easier to study than other Classes, we still know woefully little about globally threatened species (Green & Hirons 1991). Among neotropical species, less than a quarter have been subjected to any formal counting (Bibby 1994). The most common pattern is that species are inferred to be rare because they have rarely been seen. This confuses the possibilities that they are difficult to see or occur in places that are difficult of access rather than being inherently rare. In many cases, such species are then inferred to be threatened or declining because of measured (or more often qualitatively observed) habitat changes within their inferred small ranges (Stattersfield et al. 1998).

I see the collection of much better data on these species as a major challenge for bird conservationists in coming years. This would be greatly aided by quantification of what is known of range, habitats and densities. It would clarify understanding if the notions of loss of habitat and of reduction of numbers within habitats (i.e. loss of quality) could be separated (Fig. 1). Loss or conversion of habitats would best be measured by remote sensing and other large-scale approaches. Ornithologists could concentrate on estimating densities and trends within the different habitats. This approach coincides with the idea of a Natural Capital Index which has NCI as the product of quality and quantity (ten Brink 1997).

A long history of bird monitoring in the US and several west European countries has generated several findings from which environmental trends have been pointed out. Europe has mapped and reviewed the status of all breeding birds in a collaborative project involving most countries (Tucker et al. 1994; Hagemeijer & Blair 1997). These results have been taken into a European environment assessment presented to ministers (European Environment Agency 1998). These are the only quantitative measures of biodiversity trends on a wide scale included in this important report on environmental trends. Tuxill and Bright (1998) in The 1998 State of the World Report describe birds as the clearest of all indicators of biodiversity trends. These are compliments to the degree of organisation in ornithology and the extent of data on birds compared with other taxa. Results from Europe point to widespread reduction of bird populations across many species and countries. Most striking has been the reduction of once common and widespread species, especially in western Europe and primarily attributable to agricultural intensification. Because EC agriculture is so heavily directed by the Common Agricultural Policy, this is a good example of birds indicating the non-sustainability of a policy instrument. Birds have proved to be effective indicators in this context for the primary reasons that they are popular and command abundant data, largely collected by volunteers (Furness & Greenwoood 1993). No doubt similar trends could be found from amongst other taxa if only the data were available.

In spite of the impact of bird data, much attention is still needed to improve their quantity and quality. For many countries, even in Europe, the estimates of populations and trends do not stand too much scrutiny. The only overview conducted so far was able to review the period 1970–1990 and categorise species into five classes of population trend for each country (Tucker & Heath 1994). Such a formulation has proved to be powerful in influence but it cannot be regarded as being up to date and sensitive in its ability to respond to new trends.

How might ornithologists improve their data collection to make such indicators more effective and responsive? In the tropics, such data are almost completely lacking. Even the simplest review of species status by country, which initially would have to be largely qualitative would be a major advance for many countries. Some of the longer running bird monitoring schemes have been very intensive in their field work, especially those based on territory mapping (Bibby et al. 1992; Gibbons et al. 1996a). There are often sampling problems if volunteers, without any statistical design or guidance, have selected plots (Greenwood 1996). Recent thinking has considerably advanced the efficiency with which data on distribution, numbers and trends of birds can be collected (Gregory & Baillie in press; Gregory et al. in press).

Ideally, we would have adequate information on all species. At the moment, different classes of survey method are applied and they contribute differently according to the distribution and abundance of the bird (Bibby et al. 1992; Gilbert et al. 1998). Thus for very rare or concentrated species, we can aspire to good information by highly targeted surveys. Such is the situation for several species of seabirds or waterfowl which form concentrations or for some very rare species. Generic survey methods, usually designed for population monitoring, work most effectively for common and widespread species (Gilbert et al. 1998). Atlas surveys generate some information for all species (Hagemeijer et al. 1997; Harrison et al. 1997). The challenge in filling the gaps in knowledge is to move the boundaries of impact of the different survey approaches in cost-effective ways (Fig. 2).

Within the UK, the generic breeding bird survey has been moved from some 200 observer-selected plots which were surveyed by mapping to some 2000 1-km grid squares statistically selected and surveyed by transect. The costs in administrative time have fallen as a result of removing the burden of standardising the analysis of census maps. The power of the survey and the range of species for which trends can be detected has risen enormously. A similar survey started in Spain in 1996, and Hungary and the Republic of Ireland in 1998. The willingness of volunteers to contribute far exceeded expectation. There is clearly a prospect for similar work in many countries which would currently view themselves as inadequately endowed with observers.

In southern Africa, the recently published atlas (Harrison et al. 1997) has broken new ground in its scale, the impact of international collaboration and the degree of quantification which can be extracted from something as simple as the frequency of occurrence of a species on check lists. To date, bird atlas project leaders have tended to think in terms of complete coverage and very simple methods in the field (check lists). There are clearly prospects for innovative and efficient studies which design the spatial allocation of effort rather than going for full coverage. There are prospects for improving the generation of semi-quantitative data from a unit of field work (such as by collecting fixed time species lists). There are also prospects for merging atlas and generic bird monitoring thinking by designing the allocation of observer effort both across space and time.

Atlas types of survey are undoubtedly very helpful as a first contribution towards quantitatively documenting the birds of a region. On the other hand, for giving policy signals, the annual reporting of results is very important. This is partly because trends can be detected sooner and with more confidence if they are based on estimates from several years. It is also because established trends are harder to ignore if new and reinforcing data can be reported frequently.

A particular problem arises in data rich regions when it comes to summarising data on trends which have arisen from many surveys with different design and different frequencies of repeat. Such is the case in Britain and a few other countries in Europe. More species breed in Britain than at any time in recorded history. Some species such as the magpie Pica pica have increased spectacularly in numbers. Other species have declined substantially and some are very rare and vulnerable (Gibbons et al. 1996b). How do we summarise all these findings to report an overall picture of conservation progress and challenges? One possibility would be to add all the trends together and take an average. Fig. 3 suggests that this may not be very helpful. Two real examples from the UK add together to suggest very little trend over time. A most satisfactory outcome? Careful inspection shows that one species is a human commensal at the most charitable; a pest in the minds of many people. The increase of the magpie reflects the expansion of human modified habitats. The skylark is an emblematic songbird whose decline represents the retreat of biodiversity from intensively managed farmland. The difficulty is that declining skylarks and increasing magpies are not considered to counter balance each other. In the worst case, the declining species could fall extinct while the summed index, masked by the increasing species, would show no sign. Indeed the extinction of declining species could actually cause an increase as a result of removal of their negative contributions. A summed species index has formally been proposed by the UK government as one measure of sustainability. Rare and non-native species have been omitted (DETR 1998).

A novel approach to this problem has been proposed in Britain by the Royal Society for the Protection of Birds, British Trust for Ornithology and BirdLife International (in prep.). It is based on summing measures of the degree to which species population levels diverge from a target stated as a population size or index. It is proposed that the relationship between biodiversity value and population rises from zero to one but is constant at one if the population exceeds its targets (Fig. 4). Various shapes of relationship would be possible but the overall shape of the time trend in values may not be very sensitive to this parameter. It could be argued that for some species, value does not remain uniformly high above the target level but should begin to fall as they become pests. Increases in other species, while not pests, might be regarded as unsatisfactory because they are caused by adverse trends. Increasing numbers of Oystercatchers Haematopus ostralegus on Dutch farmland might be caused by rising nutrient levels which have obviously undesirable overall effect on plant and invertebrate diversity. Van Strien (1997) overcame this problem in a Dutch study by only using rare species. This seems hard to justify.

In a proof of concept paper, targets were set for 10% of Britain’s birds selected in a stratified manner across the different kinds of monitoring programmes. Ideally the targets would be formally agreed within the Biodiversity Action Plan (Anon 1995; UK Biodiversity Group 1998). Thus, they can be negotiated between stakeholders to be challenging but realistic and to reflect previous declines and plausible recovery rates over a named fixed period. Such an index could readily be assembled annually from data generated by existing monitoring schemes. Data from taxa other than birds could be added with the only constraint being the scarcity of schemes for collecting them.

A similar idea has been proposed for The Netherlands (Van Strien in press). In this case, rather than targets, it will use reference points set at recent historic population levels. Thus it will measure deviation from a point in history which may or may not be realistically recreated. It is also proposed that species will be selected for entry on the basis of representing a range of habitats. A problem with the reference year approach is that it leaves conservationists vulnerable to the charge that they are trying to turn back the clock to previous times (and by implication are opposing social development too). In many cases, data do not exist from the sufficiently distant past, so they have to be made up or estimated in some way.

Both these schemes appear to meet the key requirements of an effective indicator. They will annually report on trends in biodiversity in a quantitative manner. They will rely on the fact that it is relatively easy to design bird monitoring surveys, often using volunteers, to conduct the fieldwork. While the summary outcome is a simple index, the input data can be assembled in different ways to pick out any patterns common across habitats or likely impacts.

An important potential indicator is the extent to which ecosystems are secure. This is a Response indicator in the earlier described taxonomy. Measurement requires answers to spatial questions on the locations and qualities of ecosystems. We can then track the extent to which protected areas adequately embrace and protect the range of biodiversity or the extent to which biodiversity qualities remain outside and inside of protected areas. How good are birds as indicators of the spatial patterns of biodiversity? Not surprisingly, the answer to this question is scale dependent. It also depends on just how the question is formulated. Confusion in the literature on this subject has sometimes arisen because of confusion between biodiversity and species richness. Authors have not always asked the appropriate question for their stated purpose (McGeoch 1998).

Williams et al. (in press) present an analysis of birds and mammals in sub-Saharan Africa based on one-degree grid cells. This shows relatively poor correlation in patterns of species richness. On the other hand, if you use one data set to pick a set of grid cells representing the maximum number of species, this is quite effective for the other data set. The reason for this is that patterns of endemism are similar so a set of cells representing all species will pick up representation of endemism for different taxa.

At smaller scales, many studies (for instance Prendergast et al. 1993; Lawton et al. 1998) have shown rather poor relationship between different taxa in species richness across sampling sites. These results superficially suggest rather little prospect of using a smaller set of taxa as indicators of biodiversity (Pimm & Lawton 1998). Howard et al. (1998), using data from Uganda for moths, butterflies, birds, woody plants and small mammals also show poor correlations for species richness (having partialled out a large confounding effect of plot size). This paper goes on to show that the problem of selecting a subset (in this case 20%) of areas for protection can be solved rather effectively using just birds or butterflies. This is because there is a high degree of similarity in patterning between species formally measured as correlations between taxa in pairwise dissimilarities between sites. More simply expressed, there are a range of habitat types in Ugandan forests. A set which maximises the inclusion of species for one taxon will do so for others because it has to represent the major habitat types. If this result proves to have any generality in the humid tropics (where endemism and species richness are generally greater than elsewhere) it does allow the possibility of using a small number of taxa as indicators for this practicably important kind of spatial question. This is important because in spite of the effort invested, the surveys in Uganda were only 60–80% complete. Another study from South Africa found that grid cells chosen for different taxa were not very complementary (Van Jaarsveld et al. 1998).

This paper has tried to bring to the attention of ornithologists a fast emerging opportunity that exists for making their data politically relevant as a contribution to environmental conservation and sustainability. I argue that data on birds are well ahead of those for other taxa but ornithologists should not be complacent. In most cases, we still do not know enough in a well organised way to meet the essential requirements of environmental indicators. There is great scope to improve the collection and organisation of data on the status both of common and of rare and threatened species. There are however good recent examples of improvement in survey design and efficiency which deserve wider application.

I have argued that it will be easier to develop indicators within a planning framework that includes targets for biodiversity conservation. This will be a challenging argument to win. It involves a negotiation between scientists who can estimate what is possible at what cost and politicians concerned with the process of deciding what society is prepared to pay. Members of both disciplines might prefer not to enter such a challenging debate. For scientists, there is difficulty in extending beyond the comfort zone of reasonable certainty based on data or theory. For politicians there is the fact that accepting even simple targets such as arresting the decline of a species will turn out to have very profound effects if (as is usually the case) declines are caused by conflict with economic and sectoral interests. Few people are against conservation in general, but a target based approach makes it harder to avoid ultimate responsibility. There are some important questions to tackle about the relationship between targets, broader and longer term goals or the more neutral notion of baselines or benchmarks.

Assuming that the idea of environmental indication continues to grow, another important practical question lies just ahead. What surveys of which taxa are most appropriate to conduct and report. We should learn from the debate about spatial congruence in patterns of biodiversity here. The scientific question to tackle is not about the extent to which temporal trends in biodiversity are congruent across taxa. Rather we need to be able to identify an optimal set of taxa whose monitoring will cost effectively maximise the chances of highlighting likely future adverse trends arising from all plausible pressures. Included within the optimisation calculation is the question of cost of future data collection.

Thought needs to be given to the idea of selecting indicator species to represent particular habitats or environmental pressures. This makes sense across higher taxa if the survey methods and technical expertise required are different. One might for instance choose whether to invest in Odonata or Mollusca in adding a taxon particularly sensitive to aquatic conditions. It is harder to see why one would choose species, for instance within birds, if data on all could be collected anyway and if the choice itself had an arbitrary dimension. We simply do not know enough to predict now which species might turn out to be the most adversely impacted by an emerging phenomenon such as the cultivation of genetically modified organisms.

Ornithologists will have much to contribute to these discussions based on the wide and extensive experience of designing effective surveys and collaborative networks and on the existence of large synthesised data sets compared with those currently available for most other taxa.

Various people have helped the development of my thinking with discussions around the subject and comments on an earlier draft of this paper. I particularly appreciate the contributions of Mark Avery, David Gibbons, Cathy King, Boo Maisels, Derek Pomeroy and Arco van Strien.

Anon. 1995. Biodiversity: the UK Action Plan. London: HMSO.

Anon. 1997. Biodiversité et économie agraire: Recherche méthodologique pour la quantification et la rémunération d’un produit intentionnel en matière de conservation de la biodiversité. Brussels: Institut Royal des Sciences Naturelles de Belgique.

Bibby, C.J., Burgess, N. & Hill, D.A. 1992. Bird Census Techniques. London: Academic Press.

Bibby, C.J. 1994. Recent past and future extinctions in birds. Phil. Trans. R. Soc. Lond. B 344: 35–40.

Collar, N.J. & Andrew, A. 1988. Birds to watch: the ICBP world check-list of threatened birds. Cambridge, UK: International Council for Bird Preservation (Techn. Publ. 8).

Collar, N.J, Crosby, M.C. & Stattersfield, A.J. 1994. Birds to Watch 2: The World List of Threatened Birds. Cambridge, UK: BirdLife International.

Collar, N.J, Gonzaga, L.P, Krabbe, N., Madroño Nieto, A., Naranjo, L.G., Parker, T.A. & Wege, D.C. 1992. Threatened birds of the Americas: the ICBP/IUCN Red Data Book. Cambridge, UK: International Council for Bird Preservation.

Collar, N.J. & Stuart, S.N. 1985. Threatened birds of Africa and related islands: the ICBP/IUCN Red Data book. Cambridge, UK: International Council for Bird Preservation and International Union for the Conservation of Nature.

Department of the Environment, Transport and the Regions. 1998. Sustainability counts: consultation paper on a set of ‘headline’ indicators of sustainable development. London: DETR.

European Environment Agency. 1998. Europe’s Environment: the Second Assessment. Luxembourg and Oxford: Office for Official Publications of the European Communities and Elsevier Science Ltd.

Furness, R.W. & Greenwood, J.J.D. (eds) 1993. Birds as Monitors of Environmental Change. London: Chapman & Hall.

Gibbons, D.W., Hill, D.A. & Sutherland, W.J. 1996a. Birds. In: Sutherland, W.J. (ed.) Ecological Census Techniques a handbook. Cambridge: Cambridge University Press: 227–259.

Gibbons, D.W., Avery, M.I. & Brown, A.F. 1996b. Population trends of breeding birds in the United Kingdom since 1800. British Birds 89: 291–305.

Gilbert, G., Gibbons, D.W. & Evans, J. 1998. Bird Monitoring Methods. Sandy: Royal Society for the Protection of Birds.

Green, R.E. & Hirons, G.J.M. 1991. The relevance of population studies to the conservation of threatened birds. In: Perrins, C.M., Lebreton J.-D. & Hirons G. J. M. (eds) Bird Population Studies. Oxford: Oxford University Press: 594–633.

Greenwood, J.J.D. 1996. Basic techniques. In: Sutherland, W.J. (ed.) Ecological Census Techniques a handbook. Cambridge; Cambridge University Press: 11–110.

Gregory, R.D. & Baillie, S.R. in press. Survey design and sampling strategies for breeding bird monitoring. Proceedings of the 13th International conference of the European Bird Census Council, Estonia 1995.

Gregory, R.D., Baillie, S.R. & Bashford, R.I. in press. Monitoring breeding birds in the United Kingdom. Proceedings of the 13th International conference of the European Bird Census Council, Estonia 1995.

Hagemeijer, W.J.M. & Blair, M. (eds) 1997. The EBCC Atlas of European Breeding Birds: their distribution and abundance. London: T. & A.D. Poyser.

Hammond, A., Adriaanse A., Rodenburg E., Bryant D. & Woodward R. 1995. Environmental Indicators: A Systematic Approach to Measuring and Reporting on Environmental Policy Performance in the Context of Sustainable Development. Washington DC: World Resources Institute.

Harrison, J.A., Allen, D.G., Underhill, L.G., Herremans, M., Tree, A.J., Parker, V. & Brown, C.J. 1997. The Atlas of Southern African Birds. Johannesburg: BirdLife South Africa.

Howard, P.C., Viskanic, P., Davenport, T.R.B., Kigenyi, F.W., Baltzer, M., Dickinson, C.J., Lwanga, J.S., Matthews, R.A. & Balmford, A. 1998. Complementarity and the use of indicator groups for reserve selection in Uganda. Nature 394: 472–475.

ICBP 1992. Putting biodiversity on the map: priority areas for global conservation. Cambridge, UK: International Council for Bird Preservation.

INTERAISE 1996. World Directory of Country Environmental Studies: an Annotated Bibliography of Natural Resource Profiles, Plans, and Strategies. Washington DC: World Resources Institute.

IUCN/SSC 1994. IUCN Red List Categories. Gland, Switzerland: IUCN Species Survival Commission.

IUCN 1996. IUCN Red List of Threatened Animals. Gland, Switzerland: IUCN.

Lawton, J.H., Bignell, D.E., Bolton, B., Bloemers, G.F., Eggleton, P. Hammond, P.M., Hodda, M., Holt, R.D., Larsen, T.B., Mawdsley, N.A. & Stork, N.E. 1998. Biodiversity inventories, indicator taxa and effects of habitat modification in tropical forest. Nature 391: 72–76.

Lawton, J.H. & May R.M. (eds) 1995. Extinction Rates. Oxford: Oxford University Press.

MacGillivray, A. (ed.) 1994. Environmental Measures: Indicators for the UK Environment. London: Environment Challenge Group.

McGeoch, M.A. 1998. The selection, testing and application of terrestrial insects as bioindicators. Biol. Rev. 78: 181–201.

Miller, K.R. & Lanou, S.M. 1995. National Biodiversity Planning: Guidelines Based on Early Experiences Around the World. Washington DC, Nairobi, Gland Switzerland: World Resources Institute, United Nations Environment Programme and IUCN–The World Conservation Union.

OECD. 1993. OECD Core Set of Indicators for Environmental Performance Reviews; a synthesis report by the Group on the State of the Environment. Environmental monographs no 83, Paris: OECD.

OECD. 1996. Shaping the 21st Century: The Contribution of Development Co-operation. Paris: OECD.

Pimm, S.L. & Lawton, J.H. 1998. Planning for biodiversity. Science 279: 2068–2069.

Prendergast, J.R., Quinn, R.M., Lawton, J.H., Eversham, B.C. & Gibbons, D.W. 1993. Rare species, the coincidence of diversity hotspots and conservation strategies. Nature 365: 335–337

Secretariat of the Convention on Biological Diversity. 1997. The Biodiversity Agenda: Decisions from the Third Meeting of the Conference of the Parties to the Convention on Biological Diversity. New York and Geneva: United Nations.

Stattersfield, A.J., Crosby, M.C., Long, A.J. & Wege, D.C. 1998. Endemic Bird Areas of the World: Priorities for Biodiversity Conservation. Cambridge UK: BirdLife International.

Steadman, D.W. 1995. Prehistoric extinctions of Pacific island birds: biodiversity meets zooarchaeology. Science 267: 1123–1131.

Ten Brink, B. 1997. Biodiversity indicators for integrated environmental assessments. Bilthoven, Netherlands: UNEP/RIVM Technical Report series.

The World Bank. 1995. Monitoring Environmental Progress: a Report on Work in Progress. Washington DC: The International Bank for Reconstruction and Development/The World Bank.

The World Bank. 1997a. Expanding the Measure of Wealth. Washington DC: The International Bank for Reconstruction and Development/The World Bank.

The World Bank. 1997b. World Development Indicators. Washington DC: The International Bank for Reconstruction and Development/The World Bank.

Thirgood, S.J. & Heath, M.F. 1994. Global patterns of endemism and the conservation of biodiversity. In: Humphries, C.J. & Vane-Wright, R.I. (eds) Systematics and conservation evaluation. Oxford, Oxford University Press: 207–227.

Tucker, G.M., Heath, M.F., Tomialojc, L. & Grimmett, R.F.A. 1994. Birds in Europe: their conservation status. Cambridge, UK: BirdLife International.

Tuxill, J. & Bright, C. 1998. Losing strands in the web of life. In: Lester R Brown et al. State of the World 1998. New York: W.W. Norton and Company

UK Biodiversity Group. 1998. Tranche 2 Action Plans. Volume 1 – vertebrates and vascular plants. Peterborough, UK: English Nature

Van Jaarsveld, A.S., Freitag, S., Chown, S.L., Muller, C., Koch S., Hull, H., Bellamy, C. Kruger, M. Endrody-Younga, S., Mansell, M.W. & Scholtz, C.H. 1998. Biodiversity assessment and conservation strategies. Science 279: 2106–2108.

Van Strien, A.J. 1997. Biodiversity declining in the Netherlands: an indicator to describe the changes in the number of wild species. Neth. Official Stat. – Winter: 45–49

Van Strien, A.J. in press. From monitoring data to policy-relevant summary statistics. Bird Numbers 1998. Where Monitoring and Ecological Research meet. Proceedings 14th International Conference of EBCC.

Williams, P., Burgess, N., & Rahbek, C. in press. Hotspots of richness, hotspots of endemism, and complementary areas: their success in representing the diversity of sub-saharan mammals, their consequences for birds, and their application to large ‘flagship’ mammals for representing small mammals. In: Entwistle A. & Dunstone N. (eds) Has the Panda had its Day? Future Priorities for the Conservation of Mammalian Biodiversity. London: Chapman and Hall.

WWF/IUCN. 1994. Centres of Plant Diversity: a Guide and Strategy for their Conservation, 1: Europe, Africa, South West Asia and the Middle East. Cambridge, UK: World Wide Fund for Nature and International Union for Conservation of Nature and Natural Resources.

WWF/IUCN. 1994–1995. Centres of Plant Diversity: a Guide and Strategy for their Conservation, 2: Asia, Australasia and the Pacific. Cambridge, UK: World Wide Fund for Nature and International Union for Conservation of Nature and Natural Resources.

WWF/IUCN. 1997. Centres of Plant Diversity: a Guide and Strategy for their Conservation, 3: North America, Middle America, South America, Caribbean Islands. Cambridge, UK: World Wide Fund for Nature and International Union for Conservation of Nature and Natural Resources.

WWF. 1998. Living Planet Report 1998. Gland: WWF – World Wide Fund for Nature.

Fig. 1.  Monitoring threatened species. Half the original vegetation has been converted into a totally unsuitable habitat (C), one quarter has been modified but the bird still survives at half its original density (B) and one quarter has survived intact. Ideally, the extent of the three habitat class would be measured by remote sensing and would indicate major habitat changes. Ornithologists would estimate bird densities in the three. Bird numbers (and trends over time) are obtained as the summed products of density and habitat extent. Current knowledge is rarely adequate to codify understanding in this manner leading to confusion between observed changes in numbers and inference from changes in habitat extent.

Fig. 2.  Survey design and the development of knowledge on trends and numbers of birds. Different survey approaches work differently according to the pattern of numbers and distribution of the species. Rare or concentrated species can be counted by targeted surveys. Generic monitoring surveys work best on common and widespread species. Atlas approaches obtain some information on all species. The challenge is to improve the design and efficiency of all three classes of approach to fill the knowledge gap in the middle.

Fig. 3.  Trends in population index of two species in the United Kingdom from the Common Birds Census. With no knowledge about the species, it would be possible to conclude that these two add up to no significant trend over time – one species increases and one declines. Conservationists are not satisfied. The magpie is a human commensal increasing in ever more human impacted landscapes. The skylark is declining as intensive agriculture reduces the diversity of large areas.

Fig. 4. Possible models for the relationship between biodiversity value and population level compared with a target (which could instead be a baseline). It is postulated that the value falls as populations decline and become zero at extinction. There are subtle arguments about what shape the relationship might take. Above the target or baseline, further increases do not increase values, though it could be argued that for ‘pest’ species, they should at some stage start to fall with increasing population. Values are arbitrarily scaled to lie between 0 and 1. Annual values can be summed across species to generate a biodiversity capital index.