HBW 5 - Foreword on risk indicators and status assessment in birds by Nigel J. Collar

View foreword in PDF format: 


The Handbook of the Birds of the World (HBW) is, as reviewers never cease to indicate, an astonishing achievement, and their critical rapture has consistently been mixed with a sense of awe – almost disbelief – that such a project should have come to pass. The notion of the time being ripe for so massive an enterprise took many people, myself included, by surprise; but despite the dithering costs of such a project and the dizzying tasks involved in its management, it is certainly the case that much of the basic information on the bird species of the world has now been generated and is available for assembly. A hundred years ago HBW would have been unthinkable: so much of the planet remained to be explored, so many species remained to be discovered. Fifty years ago the picture was still very obviously incomplete; and even if new species were by then becoming much harder to find, our general biological knowledge of a high proportion of birds was abidingly feeble. The past quarter-century, however, has witnessed a radical transformation in our information base, as the growth of worldwide tourist markets and the proliferation of Peterson-style fieldguides has elevated birdwatching into a major international pursuit. The rarest birds have accordingly been pursued, their means of surefire identification frequently established, their habits and habitats broadly documented. HBW caught the wave and is riding it triumphantly to shore.

To shore, to conclusion, to an end; and, when it ends, sometime around 2010 if all goes to plan, there will lie in our hands a truly wonderful resource, a breathtaking work of reference, a masterful avian encyclopedia, a grid of information across the biosphere such as has never previously existed for any major group of animals or plants. And what else? A little less mystery, perhaps; and a lot fewer birds. I do not just mean fewer individual birds, fewer populations, and smaller ranges for almost everything except the human commensals, although this in itself is a deeply depressing prospect; I also mean fewer bird species.

There is, of course, a connexion between these things. It is an odd circumstance – everyone looks forward to the next volume of HBW, bearing, as we know it will, the most up-to-date information on a great swathe of the world's avifauna, yet most of us will also have at least an inkling that these new data owe their existence to events over which we may be decidedly less enthusiastic. Birdwatchers and biologists get to ever more remote places – islands, interiors – by virtue of new airports, new logging roads, new tourist facilities. They arrive as tiny components of the great machinery of economic development which, in a few short years, mutilates natural landscapes and human cultures beyond recognition and brings Coca-Cola, television, chainsaws, DDT and debt to every cultivable corner of the planet. By the year 2010 not only will we know more about birds than ever before (and be able to find much of that knowledge in HBW); we will also have most of them completely surrounded.

We can therefore anticipate learning very early in the new millenium the full meaning of the notion of global stewardship; if not, we will discover instead what it might be like to visit a zoo in which all of the cages are empty. So perhaps the foremost aspiration within this notion must be the absolute minimisation of extinction at the global level. There are, of course, many plans to be devised and implemented in order to staunch the steady outflow of our planet's biological diversity at the distributional, population and genetic levels – all those disappearing individuals – but this is an area riven with political and procedural complexity, and a satisfactory review of the situation, while urgently required, will take several years. Meanwhile we must make no less haste in reaching a better understanding of the early warning signs of endangerment, and a clearer appreciation of the means by which we can confidently and consistently register species as in such danger: the sooner these things are sorted, the surer we can be of preventing too great a taxic haemorrhage by the time HBW has run its spectacular course.

A  catalogue of risk indicators

We have had many years – indeed, several centuries – to accumulate the evidence; yet we still cannot be confident of all the elements that render species susceptible to extinction. Extinction probabilities appear to vary considerably from one species to another, for reasons we cannot yet – or cannot always – discern; as Simberloff (1986) put it, “Most of the traits said to make a species extinction-prone are eminently reasonable, but... the problem is that for every trait listed one can find species that have the trait and have not gone extinct.”

The most obvious and immediate explanation for this problem is that it is traits of one sort or another in combination that produce extinctions. Certainly this idea lies behind the widely used metaphor of a vortex to describe the extinction process, where one relatively innocent factor compounds, and is compounded by, other such factors to send a species into an accelerating downward spiral. The interactions of these factors are likely to become increasingly complex as the process of decline continues yet, other than in the skeletal reviews of Frankel and Soulé (1981) and Soulé (1983), I am not aware that an attempt has been made to describe even the simplest of such interactions, although population viability analysis, which models future trends on the basis of different factors at different intensities, seeks to mimic them. It does not even appear that an attempt hitherto has been made, at least as far as birds are concerned, simply to catalogue all the elements that commonly serve in the detection of extinction-proneness.

Meanwhile, since the mid-1990s we have had a widely acclaimed and approved new set of criteria for global status assessment (next section) which champions the use of quantifiable indicators, often in combination. The primary indicators (decline rate, range size and population size) are explicit; however, many secondary indicators (various types of ecological adaptation) are concealed within the general requirement that the criteria will be applied with due consideration of circumstance. The following baseline catalogue attempts to deal with both types, and is offered with three things in mind. First, it may simply be of help in identifying at an early stage the most appropriate species (on the basis of particular geographic areas, habitats, life-history strategies, and so on) for which to be most vigilant. It may also sometimes provide points of reference in cases where quantification otherwise is a matter of informed opinion and decisions to red-list depend on an arbitrary indulgence of the precautionary principle. Finally, it may act as a challenge to fieldworkers to start filling the critical gaps in our knowledge concerning the evolutionary and anthropogenic circumstances of any species suspected to be in difficulties, whether local or global.

In this essay I do not reference information taken directly from Collar and Stuart (1985), Collar et al. (1992, 1994), and HBW itself.


Small range size

For a species to become extinct, its numbers must decline and, in some measurable way, its range size must decrease. Small range size is clearly associated, albeit unpredictably, with extinction risk (see Figure 1). In a small range, any detrimental practices may so rapidly influence the entire population of a species, however abundant it may be (Terborgh 1974), that “extinction is possible within the time span which would be required for conservation measures to take effect” (de la Mare 1987). Perhaps the most notorious example of this was the 1 km2 ridge forest at Centinela near Río Palenque in Ecuador, 10% of whose flora (some 100 species) were estimated to be unique to the site, which was pristine in 1975 and totally cleared by 1988 (Dodson and Gentry 1991). The most notorious avian example may be the Stephen Island Wren Xenicus lyalli, whose range was so small that a single cat is credited with its extermination (Rothschild 1907). The need to guard against such catastrophes was precisely what motivated the BirdLife exercise to identify and map “restricted-range” species (ICBP 1992, Stattersfield et al. 1998). These are creatures that demand our eternal vigilance, most of which needs to be directed at the distribution and intensity of habitat destruction, and in particular of forest clearance. The threshold of 50,000 km2 for a restricted-range bird builds in the opportunity, in many cases, for pre-emptive conservation; conversely, the opportunity for other deleterious factors to make a serious impact escalates with diminishing range size.

Linear range

Species whose habitats are essentially linear in nature (through confinement to rivers, shorelines, etc.) are exposed to relatively elevated risks from conversion or, more usually, disruption of their habitats, and also because the one-dimensionality of their range greatly constrains total population size. Maps that generalise ranges can give a very misleading idea of the status of such species, which can appear to occupy very large areas and hence to be entirely secure. Recently it was shown that the once-abundant River Lapwing Vanellus duvaucelii is highly disadvantaged, in Laos at least, through its confinement to moderately large rivers (>30 m wide), which however are also foci for human settlement and disturbance, and are the persistent target for hydroelectric installations (Duckworth et al. 1998). Disturbance, pollution and/or siltation of rivers affect anatids like Blue Duck Hymenolaimus malacorhynchos and Brazilian Merganser Mergus octosetaceus, sandbank nesters such as Indian Skimmer Rynchops albicollis and Piping Plover Charadrius melodus, and fish-dependent forest birds like Blakiston's Eagle-owl Bubo blakistoni; it can even endanger small birds if they also occupy small ranges, such as the Rufous-throated Dipper Cinclus schulzi and Luzon Water-redstart Rhyacornis bicolor. Birds of riverine forest like the Wattled Curassow Crax globulosa suffer disproportionately if disturbance and destruction of their habitat is compounded by hunting.

Ancient isolates

Islands have been the venue for the majority of recorded avian extinctions since 1600 (King 1985). Confinement to ancient geographical features – oceanic islands, limestone caves, montane and crater lakes – represents a serious liability, owing to the fact that most of the species’s evolutionary history has been spent in situ, without external contact. Long-term isolation is in itself a defence against the customary “continental” pressures of an evolutionarily diverse environment, so adaptations such as flightlessness and tameness in island birds are traits that have been selected for in the absence of predators, but which are dramatically selected against as soon as assisted passage allows those predators (but also aggressive competitors and habitat-altering domestic stock) to enter the equation. This type of extinction-proneness has been recognised ever since the demise of the Dodo Raphus cucullatus, and the scale of the disaster that convulsed the Pacific in the wake of Polynesian man (Milberg and Tyrberg 1993, Pimm et al. 1994, Steadman 1995) has led to the notion that many avifaunas have already experienced an “extinction filter” (Balmford 1996). Even so, the horrifying loss of the Guam avifauna to the snake Boiga irregularis within the last two decades (Savidge 1987) demonstrates that there are still many constitutively defenceless species for which vigilance is needed. These are, however, very difficult cases: no matter how great the precautions, almost certainly something – very possibly disease (Diamond 1984a, Ralph and van Riper 1985) – will eventually slip through the barrier. (It is worth noting that the problem of long isolation extends to Australia, devastated as it is by rabbits, cats and foxes, with birds like the Gouldian Finch Chloebia gouldiae and Star Finch Neochmia ruficauda suffering from exotic invaders in the form of a parasitic mite and a habitat-smothering vine respectively.)

Political isolates

Confinement to one or two countries is a non-biological circumstance that nevertheless counts as a form of constitutive liability, rendering species vulnerable to the unpredictability of the political or economic environment. Civil unrest in Peru made it extremely difficult throughout the early 1990s to contemplate any intervention
– external or internal – on behalf of the Junin Flightless Grebe Podiceps taczanowskii (a Critically Endangered ancient isolate). The persistent war within Angola has had the same effect on the rarest Scarp endemics (six listed as Endangered in Birds to watch 2). U.S. policy towards Cuba cuts off important potential funding for the conservation of the many threatened birds there, nine of which are unique to the country. No-one can get into northern Iraq to discover the status of bustards, or into southern Myanmar (Burma) to find out if, after over a decade of intense effort for the “last population” in Thailand, Gurney's Pitta Pitta gurneyi survives in the Tenasserim tracts. Cambodia, best hope for the Giant Ibis Pseudibis gigantea, continues to make virtual bystanders of aid workers and environmentalists. Moreover, when there is serious humanitarian work to do, even the most ardent conservationist is unlikely to feel like knocking on ministry doors; if refugees are camped on the type and only locality of the Somali Long-clawed Lark Heteromirafra archeri, then peace be with them. In any case, as everyone knows who has made a long habit of knocking on ministry doors, the influence of corruption in governmental organisations – unenviably underfunded while charged with regulating the exploitation of often enormously valuable resources – is all-pervasive; indeed, the role of political corruption in the erasure of biodiversity is one of those ultimate causes of extinction that cries out for exposure and analysis.


Small population size

Risk of extinction increases with decreasing population size (see Figure 2). Among the forces that conspire to put small populations at increased risk are demographic stochasticity (chance events like an entire generation being the same sex), environmental stochasticity (chance events like storms and fires), genetic deterioration and social dysfunction (Simberloff 1986). The “effective” (roughly meaning breeding) population size is generally used, but many populations possess barely detectable and hence unquantifiable numbers of “ floaters”, whose value as rapid substitute breeders cannot be ignored. The debate about what constitutes a minimum viable population size continues to evolve (Franklin and Frankham 1998a,b Lynch and Lande 1998), but it is apparent that some species suffer no great harm from passing through genetic bottlenecks, at least as long as the bottleneck is brief (so that most genetic variation is retained): the Seychelles Warbler Acrocephalus sechellensis hit a low of c.50 in 1965 (2,060 in 1997), and the Chatham Islands Robin Petroica traversi was down to five in 1980 (250 in 1998) (recent information per A. J. Stattersfield).

Population decline

Decline is a clear indicator of adversity (again see Figure 2), and both measuring and explaining decline are crucial to an understanding of the nature and mitigation of the problem. A great challenge for conservation assessment is simply to detect significant declines, which can be particularly difficult for species with large ranges, since early status reports (often uncritically repeated, so that they seem relatively recent) can mask modern circumstances. The Spot-billed Pelican Pelecanus philippensis and Greater Adjutant Leptoptilos dubius once had huge ranges and populations (the former purportedly in millions) focused on Myanmar (Burma), but in the past century these ranges have been “hollowed out”, leaving the pelican moderately secure only in Sri Lanka and the stork at the brink of extinction. Yet neither bird was in the second edition of the international Red Data Book (King 1978-1979); nor – in spite of Stresemann and Grote's (1943) concern, which, alas, found expression in the wrong language at the wrong moment (back to political isolates!) – was the now unfindable Slender-billed Curlew Numenius tenuirostris. These birds will not have died off – or died back – entirely in vain if only they serve as warnings to maintain the monitoring and red-listing of free-falling species like the Lesser Kestrel Falco naumanni and Marbled Teal Marmaronetta angustirostris until such time as the declines are understood and tempered. Most of all they serve simply to indicate that the regular gathering of good-quality information on the status of all species is essential to detect declines before they take on terminal characteristics.


The number and size of subpopulations (difficult to judge though they often are), in particular the size and condition of the largest of them, may have a direct bearing on the fate of species and their need for management. Part of the concern is that highly sedentary species will end up in a series of essentially independent subpopulations, none of which is sufficiently large to avoid long-term effects of inbreeding (some Himalayan galliforms such as Cheer Pheasant Catreus wallichii and Blyth's Tragopan Tragopan blythii are perhaps strong avian examples). Moreover, modelling of sink/ source dynamics shows that, even in more vagile species, the majority of individuals can sometimes reside in non-sustaining subpopulations (Howe et al. 1991). Of course the relative productivity of different subpopulations can only be discovered through long-term study of breeding success and survival, but a body of evidence is emerging that elevated predator incursion, diminished food supply and other factors conspire to render many small fragments essentially unproductive (Kattan et al. 1994, Turner 1996, Robinson 1998). Certain guilds of bird appear to suffer disproportionately from fragmentation, including large raptors, large canopy frugivores and large terrestrial or near-terrestrial insectivores (Terborgh and Winter 1980, Diamond 1984b).


Habitat choice

A species's choice of habitat determines its fate. Habitat destruction is now the most important cause of extinction, responsible for 50% of continental and 20% of island bird losses (Diamond 1984a), and it is much the greatest cause of bird endangerment (Collar et al. 1994, 1997). However, habitats are distributed unevenly and differ in the types and levels of human pressure they endure, so the way that measures of habitat are used as indicating susceptibility will vary widely. Simply the rate of decline in some more circumscribed habitats, and simply the extent of others, is evidence enough to red-list some species, on the basis of traits already discussed. The phenomenon of extreme lowland rainforest specialism described by Wells (1985) in Malesia predicates whole suites of species with special needs. Natural patchiness may reflect camouflaged historical or ecological processes (Diamond 1980) and may in turn be camouflaged by generalised range maps, leading (as indicated above) to serious misjudgements of status. Confinement to habitat dominated by single plant species is usually better expressed as a feeding or breeding specialisation (see below), but Kirtland's Warbler Dendroica kirtlandii, with its remarkable summer requirement of stands of 8–20-year-old Jack Pines Pinus banksiana, is perhaps an exception. Habitats can, of course, be optimal, suboptimal or marginal, and their relative values can only really be identified by long-term studies of breeding success and survival rates. (Perhaps it should also be noted here that over the next century global warming can be expected to shrink many already severely reduced habitats, causing them and their species to wink out in ways over which we are, on present form, unlikely to have the slightest control.)

Indirect upshots of habitat destruction include (a) the breakdown of barriers that keep two species apart, resulting in hybridisation and, in some cases, genetic swamping of one species by the other (Australia's Black-eared Miner Manorina melanotis is in this kind of trouble); and (b) the disproportionate benefit it gives to certain brood parasites, resulting in small local populations of hosts suffering total reproductive failure (the Brown-headed Cowbird Molothrus ater now jeopardises three species, while the Shiny Cowbird M. bonariensis does the same, actually or potentially, for several more).

Habitat sensitivity

Complex stable communities, which generally involve a high proportion of sedentary “equilibrium” species (K-selected; see below), tend to simplify under stress. This is one of the great take-home messages of ecology: if a pristine ecosystem is disturbed, for example by hunting or by logging, then some components of that ecosystem will suffer (Erwin 1988, Brown and Brown 1992, Johns 1997, Whitmore 1997). The widespread view that natural resources must provide economic benefits in order to win vital conservation support is often expressed in the compelling, rebarbative phrase “ use it or lose it”; but the truth is that to use it is to lose it – some of it, anyway. The sensitivity of certain bird species –  notably terrestrial and lower arboreal insectivores –  to habitat modification (e.g. Lambert 1992, Thiollay 1992, Marsden 1998) needs systematic documentation, and the equilibrium of their environments must be strictly maintained: for example, protected areas within their ranges will have to be managed in such a way that large cores are kept free of human perturbation. The phenomenon of trophic simplification has been shown on land and in water; notably the loss of top predators destabilises communities by allowing the increase of smaller predators, thus causing surges and drops in numbers further down the food chain, and shaking out a number of species altogether (Terborgh and Winter 1980, Lowe-McConnell 1987, Soulé et al. 1988).


Migration increases the number of areas, habitats and food resources that a species depends on, and while, in the absence of anthropogenic influences, this predicates no greater inherent vulnerability, the risks certainly multiply when the global environment is patchily deteriorating. Migration therefore also creates a much more complex information requirement, relating to each site and stage of the annual or life cycle. Radio-tracking of Resplendent Quetzals Pharomachrus mocinno and other Central American montane bird species has demonstrated their unexpectedly complex patterns of habitat dependence through the year (e.g. Powell and Bjork 1994), and these findings are matched by the discovery of different habitats and strategies used by male, female and juvenile warblers over the non-breeding period (Hutto 1998). Although the problem is clearly most acute for species with already highly circumscribed ranges, even very widespread species can be compromised; I strongly suspect that the loss of the Slender-billed Curlew is attributable to loss of its staging grounds – presumably the Russian steppes – between summer and winter quarters, and equally I guess that the Eskimo Curlew Numenius borealis failed to recover its numbers owing to the obliteration of the prairies, which it was known to use on spring stopover. Moreover, where species form seasonal concentrations, as the Eskimo Curlew certainly did, they elevate their susceptibility by temporarily reducing their range size (see above); and where they move to areas inaccessible for political reasons, they jeopardise all the endeavours made elsewhere on their behalf.

Feeding specialisation

Feeding specialisation is a high-risk/high-return strategy. The return is that competition for the resource is minimal; the risk is that disappearance of the resource is terminal. Mistletoe specialists, like White-cheeked Cotingas Zaratornis stresemanni and Painted Honeyeaters Grantiella picta, and fig specialists, like Large Green-pigeon Treron capellei (Lambert 1991; see Figure 3) and Pesquet's Parrot Psittrichas fulgidus (Mack and Wright 1998), face serious consequences if particular trees or particular tracts are cleared. Too singular a dependence on one species of palm may lie behind the disappearance of the Glaucous Macaw Anodorhynchus glaucus, and it certainly lies behind the critical imperilment of Lear's A. leari; indeed, any strong dependence on fruit or nectar represents a risk when local climate and habitat loss conspire to isolate these resources in both time and space (a fragmentation issue). The Kaka Nestor meridionalis needs insect-derived honeydew to bring it into breeding condition, but introduced wasps also favour this resource and are getting to it first. Rats may have removed the prey-base of the presumed extinct Snail-eating Coua Coua delalandei, and medieval man certainly did so with the Canarian Black Oystercatcher Haematopus meadewaldoi. A single pair of Siberian Cranes Grus leucogeranus needs miles of narrow tundra lakeshore vegetation in which to find the food to rear its single offspring (Potapov 1992); this discovery illuminates the difficulty such a K-selected species has in recovering from losses elsewhere in its range.


Nomadism is a distinctive form of food specialisation in which a species is liberated (or driven) from a geographical position for all or part of the annual cycle and then ranges erratically within often very wide limits. Such unpredictability renders conservation planning highly problematic. The Passenger Pigeon Ectopistes migratorius was a specialist on a food (mast) that was patchy in both space and time, and which therefore needed social facilitation (strung-out fronts of flying birds) to find sufficient resources to feed and breed: despite the incredible slaughter the species endured, habitat destruction appears to be primarily responsible for its loss, since the patchier the food became, the more difficult it was for individuals to find it (Bucher 1992). Characteristics of nomadic specialisation include (a) high mobility, (b) high investment in social facilitation, (c) some degree of food specialisation, and (d) breeding opportunism. Amongst threatened birds that exhibit some of these traits are the Thick-billed Parrot Rhynchopsitta pachyrhyncha (a pine-seed specialist), Golden-plumed Parakeet Leptosittaca branickii (perhaps heavily dependent on Podocarpus fruits), Purple-winged Ground-dove Claravis godefrida and various Sporophila (bamboo specialists, as perhaps was Bachman's Warbler Vermivora bachmani).

Breeding specialisation

Hole-nesters tend to be forest and woodland climax species, making use of mature, senescent and dead trees large enough and (at least in parts) rotten enough to house them over the breeding cycle. Large-bodied birds like hornbills necessarily require massive trees; but in parts of Australia even moderate-sized species depend on extremely slow-growing eucalypts, some of which are believed to take half a millenium to reach an age at which appropriately deep splits and fissures begin to appear in their upper trunks and branches (Mawson and Long 1994). Studies on the island of Sumba in Indonesia have shown that the hole-nesting avifauna of a forest tract may entirely depend on a small number of particularly large trees, often figs (Marsden 1995). Forestry management regimes which fail to take account of these facts stand to wipe out entire groups of species. Obligate bank-nesters are also constrained: the Blue Swallow Hirundo atrocaerulea utilises Aardvark Orycteropus afer burrows and potholes, while the Three-toed Jacamar Jacamaralcyon tridactyla needs stream-hewn banks in forest remnants: both species are uncomfortably constrained by these circumstances. White-necked and Grey-necked Picathartes Picathartes gymnocephalus and P. oreas need boulders and rockfaces inside primary forest on which to build their nests, which are consequently easily and repeatedly plundered by human nest-despoilers. The Edible-nest Swiftlet Aerodramus fuciphagus suffers the same problem, and indeed this, along with other cave-nesting swifts and the Oilbird Steatornis caripensis, share with many seabirds and some megapodes, including the threatened Maleo Macrocephalon maleo, the inherent susceptibility of being obliged by geology and geography to form site-specific seasonal concentrations (i.e. a temporarily small range size).


The intrinsic rate of increase (r) in species affects their ability to recover; those with low reproductive rates – called “ K-selected” since they generally occur in highly stable environments whose carrying capacity (K) is permanently at its upper limit
– are notably susceptible to depletion by human exploitation, typically driven by maximum profit targets without regard to the long term. Greatly deferred maturity, protracted generation length, high life expectancy and small clutch-size all predict susceptibility. Albatrosses are classic examples, and the loss of only relatively small numbers of some species to the tuna industry's longlines is causing an inexorable population decline, as rates of recruitment fail to keep pace with rates of depletion (Croxall and Gales 1998). Many other seabirds, as well as some of the larger raptors, storks, ibises and cranes, exhibit similar life-history traits that may disadvantage them in the face of direct and indirect human influences.

Economic attributes

Certain kinds of species possess characteristics whose attractiveness to man is clearly disadvantageous. Birds suffer from human over-exploitation for food, for plumes and other body-parts, and for some form of companionship. The key attributes are edibility, visual or vocal beauty, and ease of capture (sometimes also ease of maintenance); cultural and medicinal properties do not seriously figure. Of course edibility is a common condition in birds, but some species are obviously more worth pursuing than others: among the more usual targets are (a) large-bodied, easily snared species of the forest floor, and (b) sociable species such as nest in colonies (especially those tied to specific sites, which simplifies access to their eggs as well), roost in flocks, and come to decoys. For some people, inevitably, the beauty of birds is the price they can charge for them, but at least plume-hunting is now commercially extinct and the extensive hunting of birds for ceremonial adornment is largely confined to parts of the Amazon (macaws and other parrots), the Greater Sundas (hornbills) and New Guinea (birds of paradise, parrots and others); possibly the only taxa seriously endangered by traditional use were in the Pacific – Vini lorikeets and certain Hawaiian honeycreepers. However, the cagebird trade remains a major cause of endangerment for some birds, although mainly only when in combination with habitat destruction. Vegetarian species are hugely preferred, for reasons of maintenance; within this subset, parrots are prime targets because of their colours, mimicry and domesticity. Such a combination of attributes is, however, rare in other major groups of species, many of which suffer from intense local rather than international interest (Yellow Cardinal Gubernatrix cristata in Argentina, Green Avadavat Amandava formosa in India, Straw-headed Bulbul Pycnonotus zeylanicus and Java Sparrow Padda oryzivora in Indonesia).

A system for status assessment

In the 1990s the International Union for Conservation of Nature and Natural Resources (now known as IUCN–The World Conservation Union) established formal criteria for use in evaluating the probability of a taxon becoming extinct (IUCN Species Survival Commission 1994). In the course of 1993– 1994, while the criteria were still being finalised, BirdLife used them to determine the species in Birds to watch 2 (Collar et al. 1994). Thus the first and second volumes of HBW reflected BirdLife/ICBP use of the old IUCN criteria, as applied in the most recent Red Data Books (Collar and Stuart 1985, Collar et al. 1992) and the first Birds to watch (Collar and Andrew 1988) –  this last requiring a degree of interpretation for the Asian and Pacific species since it did not attempt to allocate categories to species. Since 1994 HBW has been able to draw on Birds to watch 2, and from 2000 there will be a further and much more detailed revision which will serve thereafter as the point of reference for species status in the HBW accounts.

The old IUCN criteria for classifying threatened species consisted of five comfortably vague (but, with common sense, perfectly serviceable) status categories, two of which explicitly indicated probability of extinction and three of which did not. These have now been replaced with a much more demanding (but equally serviceable) set of criteria, subcriteria and qualifiers, all of which are related to three categories of threat expressing different probabilities of extinction within particular time-frames. The huge advantage of this new system is that it forces a degree of translucency and consistency into a process which previously offered shelter from these things (and therefore freedom from public accountability, something which ran the risk of bringing discredit to the entire red-listing enterprise).

The new criteria depend on the user being able to make some credible estimate of numerical circumstance. As noted earlier, a species can only become extinct by losing its numbers and shrinking some aspect of its range, and these are the two obvious things to measure when seeking to evaluate whether the probability of extinction is increasing for a taxon or not. So the five measurements available are: (1) high rate of population decline irrespective of overall numbers; (2) significant rate of population decline linked to range size; (3) significant rate of decline in range linked to population size; (4) three stand-alone population thresholds and a single stand-alone range threshold; and (5) a population viability analysis which predicts extinction within given time-frames.

The new IUCN criteria

After Birds to watch 2 went to press, the IUCN criteria were very slightly altered; and they will almost certainly be altered – adjusted – again in the near future. What follows is a brief account of the criteria as they currently stand, but fuller details can be found in the official booklet (IUCN SSC 1994) or at the following websites:





The categories and criteria below are reproduced almost wholly verbatim from IUCN SSC (1994); they are also summarised in Table 1. Several definitions needed to interpret the criteria are appended at the end.

Critically Endangered

A taxon is Critically Endangered (CR) when it is facing an extremely high risk of extinction in the wild in the immediate future, as defined by any of the following criteria (A to E):

A   population reduction in the form of either of

(1)    an observed, estimated, inferred or suspected reduction of at least 80% over the last 10 years or 3 generations, whichever is the longer, based on (and specifying) any of: (a) direct observation; (b) an index of abundance appropriate for the taxon; (c) a decline in area of occupancy, extent of occurrence and/or quality of habitat; (d) actual or potential levels of exploitation; (e) the effects of introduced taxa, hybridisation, pathogens, pollutants, competitors or parasites;

(2)    a reduction of at least 80%, projected or suspected to be met within the next 10 years or 3 generations, whichever is the longer, based on (and specifying) any of (b), (c), (d) or (e) above;

B   extent of occurrence estimated to be less than 100 km2 or area of occupancy estimated to be less than 10 km2, and estimates indicating any two of:

(1)    severely fragmented or known to exist at only a single location;

(2)    continuing decline, observed, inferred or projected, in any of: (a) extent of occurrence; (b) area of occupancy; (c) area, extent and/or quality of habitat; (d) number of locations or subpopulations; (e) number of mature individuals;

(3)    extreme fluctuations in any of: (a) extent of occurrence; (b) area of occupancy; (c) number of locations or subpopulations; (d) number of mature individuals;

C   population estimated to number less than 250 mature individuals and either:

(1)    an estimated continuing decline of at least 25% within 3 years or 1 generation, whichever is longer or

(2)    a continuing decline, observed, projected, or inferred, in numbers of mature individuals and population structure in the form of either: (a) severely fragmented (i.e. no subpopulation estimated to contain more than 50 mature individuals); (b) all individuals in a single subpopulation;

D   population estimated to number less than 50 mature individuals;

E    quantitative analysis showing the probability of extinction in the wild is at least 50% within 10 years or 3 generations, whichever is the longer.


A taxon is Endangered (EN) when it is not Critically Endangered but is facing a very high risk of extinction in the wild in the near future, as defined by any of the following criteria (A to E):

A   population reduction in the form of either of:

(1)    an observed, estimated, inferred or suspected reduction of at least 50% over the last 10 years or 3 generations, whichever is the longer, based on (and specifying) any of: (a) direct observation; (b) an index of abundance appropriate for the taxon; (c) a decline in area of occupancy, extent of occurrence and/or quality of habitat; (d) actual or potential levels of exploitation; (e) the effects of introduced taxa, hybridisation, pathogens, pollutants, competitors or parasites;

(2)    a reduction of at least 50%, projected or suspected to be met within the next 10 years or 3 generations, whichever is the longer, based on (and specifying) any of (b), (c), (d), or (e) above;

B   extent of occurrence estimated to be less than 5,000 km2 or area of occupancy estimated to be less than 500 km2, and estimates indicating any two of:

(1)    severely fragmented or known to exist at no more than 5 locations;

(2)        continuing decline, inferred, observed or projected, in any of: (a) extent of occurrence; (b) area of occupancy; (c) area, extent and/or quality of habitat; (d) number of locations or subpopulations; (e) number of mature individuals;

(3)    extreme fluctuations in any of: (a) extent of occurrence; (b) area of occupancy; (c) number of locations or subpopulations; (d) number of mature individuals;

C   population estimated to number less than 2,500 mature individuals and either:

(1)    an estimated continuing decline of at least 20% within 5 years or 2 generations, whichever is longer, or

(2)    a continuing decline, observed, projected, or inferred, in numbers of mature individuals and population structure in the form of either: (a) severely fragmented (i.e. no subpopulation estimated to contain more than 250 mature individuals); (b) all individuals in a single subpopulation;

D   population estimated to number less than 250 mature individuals;

E    quantitative analysis showing the probability of extinction in the wild is at least 20% within 20 years or 5 generations, whichever is the longer.


A taxon is Vulnerable (VU) when it is not Critically Endangered or Endangered but is facing a high risk of extinction in the wild in the medium-term future, as defined by any of the following criteria (A to E):

A   population reduction in the form of either of:

(1)    an observed, estimated, inferred or suspected reduction of at least 20% over the last 10 years or 3 generations, whichever is the longer, based on (and specifying) any of: (a) direct observation; (b) an index of abundance appropriate for the taxon; (c) a decline in area of occupancy, extent of occurrence and/or quality of habitat; (d) actual or potential levels of exploitation; (e) the effects of introduced taxa, hybridisation, pathogens, pollutants, competitors or parasites;

(2)    a reduction of at least 20%, projected or suspected to be met within the next 10 years or 3 generations, whichever is the longer, based on (and specifying) any of (b), (c), (d) or (e) above;

B   extent of occurrence estimated to be less than 20,000 km2 or area of occupancy estimated to be less than 2,000 km2, and estimates indicating any two of:

(1)    severely fragmented or known to exist at no more than 10 locations;

(2)    continuing decline, inferred, observed or projected, in any of: (a) extent of occurrence; (b) area of occupancy; (c) area, extent and/or quality of habitat; (d) number of locations or subpopulations; (e) number of mature individuals;

(3)    extreme fluctuations in any of: (a) extent of occurrence; (b) area of occupancy; (c) number of locations or subpopulations; (d) number of mature individuals;

C   population estimated to number less than 10,000 mature individuals and either:

(1)    an estimated continuing decline of at least 10% within 10 years or 3 generations, whichever is longer, or

(2)    a continuing decline, observed, projected, or inferred, in numbers of mature individuals and population structure in the form of either: (a) severely fragmented (i.e. no subpopulation estimated to contain more than 1,000 mature individuals); (b) all individuals in a single subpopulation;

D   population very small or restricted in the form of either of:

(1)    population estimated to number less than 1,000 mature individuals;

(2)    population characterised by an acute restriction in its area of occupancy (typically less than 100 km2) or in the number of locations (typically less than 5);

E    quantitative analysis showing the probability of extinction in the wild is at least 10% within 100 years.

Lower Risk

A taxon is Lower Risk (LR) when it has been evaluated and does not satisfy the criteria for any of the categories Critically Endangered, Endangered or Vulnerable. Taxa included in the Lower Risk category can be separated into three subcategories: (1)Conservation Dependent (cd) – taxa which are the focus of a continuing taxon-specific or habitat-specific conservation programme targeted towards the taxon in question, the cessation of which would result in the taxon qualifying for one of the threatened categories above within a period of 5 years; (2) Near Threatened (nt)
– taxa which do not qualify for Conservation Dependent, but which are close to qualifying for Vulnerable; (3) Least Concern (lc) – taxa which do not qualify for Conservation Dependent or Near Threatened.

Data Deficient

A taxon is Data Deficient (DD) when there is inadequate information to make a direct, or indirect, assessment of its risk of extinction based on its distribution and/or population status. A taxon in this category may be well studied, and its biology well known, but appropriate data on abundance and/or distribution are lacking. Data Deficient is therefore not a category of threat or Lower Risk. Listing of taxa in this category indicates that more information is required, and acknowledges the possibility that future research will show that threatened classification is appropriate.

Not Evaluated

A taxon is Not Evaluated (NE) when it is has not yet been assessed against the criteria.

Further definitions

The following terms require definition in relation to the above: population, the total number of individuals of the taxon (expressed as mature individuals only); subpopulation, geographically or otherwise distinct groups in the population between which there is little genetic exchange; generation, the average age of parents in the population; extent of occurrence, the area contained within the shortest continuous imaginary boundary which can be drawn to encompass all the known, inferred or projected sites of present occurrence of a taxon; area of occupancy, the area within its extent of occurrence which is occupied by a taxon (usually only measurable in terms of the area of its habitat); location, a geographically or ecologically distinct area in which a single event will soon affect all individuals of the taxon present.

The problems of data quality and consistency of judgement

The new IUCN criteria are analogous to a piece of legislation: no matter how clear, objective and straightforward they are intended to be, the options and qualifications they possess, in their attempt to cater for different biological and heuristic circumstances, mean that their consistent application using anything less than high-quality data is unattainable. Minor loopholes and grey areas allow scope for various interpretations and hence result in variable assessments (Stattersfield 1997).

The fundamental problem is that, despite the ornithological knowledge base being so much broader than it was 25 years ago (enough to write a Handbook entry on every bird), it is still insufficient to sustain confident judgements on the conservation status of a significant minority of tropical species. It is not just that there are very few species for which the exact number of individual specimens alive is known, or the exact range. It is also, far more problematically, that assessments of the severity of threats to species, most obvious among them being rates of habitat loss, are commonly (a) out of date, (b) difficult to interpret, (c) incomplete, (d) unquantified, (e) patently unreliable – and any of the foregoing in combination – or otherwise (f) non-existent. Writing this at the start of 1999, I still have no clear idea of what happened in Borneo in the recent notorious fires of 1997–1998, yet the inclusion or exclusion of several Bornean endemic bird species from Threatened birds of Asia hangs on some understanding of the resultant status of the island's once-extensive forests.

Aside from the birds, which are all reviewed by staff at BirdLife International working with bird specialist groups, species status assessments are generally undertaken by individual specialists on particular groups of animals and plants; but these people are not, and cannot be expected to be, specialists on the status of their species' habitats. It is inevitable that different evaluators will use different background information (not only on threats but also on conservation measures) owing to differences (a) in the accessibility of resource material, (b) in the diligence with which they seek it, and (c) in the interpretation they place on it. For example, in a circumstance where four primary-forest species have the same range, the granting of a logging concession within that range might produce the following responses among four different evaluators:

•  A assumes that the concession spells disaster, and classifies CR;

•  B believes that it will take a decade to implement, and classifies EN;

•  C knows that it is on such difficult terrain that extraction will be modest, and classifies VU; while

•  D misses the story altogether, and classifies LR.

In a similar way, differences in panic and information levels over the recent fires in Borneo will doubtless cause equivalent disparities amongst the multiplicity of specialists assessing the conservation status of the island's endemics.

Background (i.e. non-specialist) information can thus be crucial to the red-listing of a species. Moreover, an even treatment of all taxa – i.e. between evaluators – is essential if there is then to be an equitable distribution of resources to support those taxa as a consequence of their red-listing: it is clearly unsatisfactory that misclassifications based on negligent research can qualify undeserving taxa for – and disqualify deserving taxa from – conservation support. No less importantly, therefore – and this is one of the main reasons for and advantages of Red Data Books as against Red Lists – the background information used in an evaluation needs to be stated, otherwise the entire process remains opaque. This is not just a matter of honesty, for if all that is published is the notation (e.g. VU C2b, a single declining population of under 10,000), then it can easily be assumed by later workers that the population figure is based on hard fact when all it represents is a highly precautionary inference using very slight evidence (Stattersfield 1997).

The preamble to the new IUCN system encourages evaluators who find themselves in a data vacuum to “make intelligent inferences” about the taxon they are assessing, and to “apply the precautionary principle” in a credible manner. In practice, of course, the degree of variation in actual or notional estimates relating to the criteria can be so great that what is “ intelligent” or “credible” is impossible to decide. The Birds to watch 2 team coined the term “responsible pessimism” in an attempt to deny the use of the worst-case scenario in making a decision, and to reduce default use of the category “ Data Deficient”. Even so, there is always the haunting fear that to go for any more comforting scenario is to risk both complacency and opprobrium, and I strongly suspect that this is a common experience among evaluators, who would far rather be ridiculed for listing a safe species than be vilified for omitting a struggling one.

Perhaps the most problematic aspect of the classification process lies in the significance to be attached to human intervention in its various forms. The category of “Conservation Dependent” was introduced because the degree of active protection a species enjoys cannot be ignored as a determinant of its status, yet the actions taken may not be adequate or appropriate, and even if they are they may not remain so. Governments cut funding to programmes; protected areas are degazetted or simply overrun; projects mistake the causes of the problem they are tackling. In addition, there is real difficulty in assessing the overall relevance and impact of intervention when it affects only relatively small parts of the total range or population of the species; and, for example, over whether species whose ranges are wholly encompassed by protected areas, and which face no threat within them, merit listing at all. These are persistent difficulties which can only be resolved on a case-by-case basis, yet for which consistency of judgement between cases is vital.


All risk indicators help identify threatened species, but those that are quantifiable are in a sense more relevant: these are the primary indicators that have been accessed as criteria in the new IUCN system. The ecological attributes are secondary indicators which, by not being measurable in space (ranges) or by numbers (populations), remain essentially more speculative and anecdotal. Nevertheless they have a distinct value in pointing to the kind of species to be considered for red-listing, and in justifying a final decision on categorisation in the absence of more formal numerical data (see Figure 3).

Growth in the composition of Red Lists provides some evidence for the way in which the world is coming to an end – perhaps not a literal end, but the end of an era to be sure. The year 2000 is a ridgetop from which the planet's lovers of wild places can look back at a rich landscape haunted by wasted opportunities, and peer forward at a dustbowl signposted with wasted words, “biodiversity” prime among them. But there can be no turning back, and indeed the next few decades – which means, pretty much, the rest of our lives – will be decisive for the wildlife of the planet for the next few millenia. So if we want the decidedly crumbly future to be worth the walk, we have to set about the pursuit of information and, in direct consequence, the pursuit of conservation, with an altogether new and messianic intensity. There is no reason why this needs to be left to professionals; amateurs can (and do) make a huge difference. The latitude necessarily given over to guesswork and assumption in IUCN status assessments is far too great, but even for the species that are not red-listed there are gaping holes in our knowledge which we cannot afford to leave unaddressed for much longer.

Naturally enough it is the threatened species on which we need to focus the maximum effort in generating new data and in implementing sound measures. The BirdLife Red Data Books have been written for people seeking to help with these species, and over the years many fieldworkers, amateur and professional, have taken up individual cases and groups of cases at given sites. For all the other species there is no better single guide to the general state of our knowledge than the Handbook of Birds of the World. Many of the primary indicators are discernible in the status sections provided and from the maps, and often the secondary indicators – feeding and breeding specialisations, seasonality, nomadism, demography and exploitation levels – are also accounted for in such a way as to be helpful to the evaluator. So in spite of my disquiet at the means by which our knowledge will be obtained, of course I welcome each new volume of HBW, and I look forward to the time when it becomes the hard-copy database for all the world's birds. If that database duly grows broader as a consequence of the individual efforts of HBW's purchasers and readers, then we shall have still more reason to be grateful.

However, the single most important perception that follows from the criteria and, in particular, the catalogue is that the majority of extinction-prone species can only be secured by protected areas, many of them large, many of them strict; only, in other words, by setting aside significant tracts of the planet with the full intention that the factors rendering their inhabitants extinction-prone shall be absolutely minimised. It is futile to pretend that this can always happen, of course, and in already “over-developed” regions like western Europe there are inevitable compromises to be made over multiple uses of the landscape. Equally, however, it is futile to pretend that man and biodiversity are more than modestly compatible bedfellows, for the fate of the great majority of species is directly related to the levels of disruption and destruction inherent in modern human economic activity. Biodiversity is like the Lascaux cave art: the more contact we have with it, the quicker it dissolves away (use it: lose it). So if conservationists, planners, development agents and politicians – the people who most often invoke the term “biodiversity” – really want to honour its meaning, then they must understand that their primary responsibility is to defend key areas –  and, I repeat, large areas –  of land and sea from the gross intrusions of the species to which they belong.

Not, of course, that even this will be enough. For protected areas to survive for more than a generation or two, a genuine revolution is needed in human socio-economic activity and relations, involving among other things the integration of conservation interests and practices across even the least natural, most brutalised landscapes. It is an awesome prospect, but one we cannot shrink from: if the green agenda seems never-ending, this is only because we want life on earth to be that way, too.


Special thanks go to A. J. Stattersfield for her help with this essay, to A. J. Balmford and T. M. Brooks for their comments on the submitted draft, and to C. J. Bibby, M. I. Evans, K. J. Gaston, J. R. Ginsberg, B. Heredia, M. G. Kelsey, N. Leader-Williams, G. M. Mace, E. J. Milner-Gulland, S. N. Stuart, G. M. Tucker, D. C. Wege and S. M. Wells for their comments on an early draft.


N. J. Collar


Balmford, A. (1996). Extinction filters and current resilience: the significance of past selection pressures for conservation biology. Trends Ecol. Evol. 11: 193-196.
BirdLife International (in prep.). Threatened Birds of the World. BirdLife International, Cambridge, U.K.
Brown, K.S. & Brown, G.G. (1992). Habitat alteration and species loss in Brazilian forests. Pp.119-142 in T.C. Whitmore & J.A. Sayer, eds. (1992). Tropical Deforestation and Species Extinction. Chapman & Hall, London.
Bucher, E.H. (1992). The causes of extinction of the Passenger Pigeon. Pp.1-36 in D.M. Power, ed. (1992). Current Ornithology 9. Plenum Press, New York.
Collar, N.J. & Andrew, P. (1988). Birds to Watch: the ICBP World List of Threatened Birds. Techn. Publ. 8, International Council for Bird Preservation, Cambridge, U.K.
Collar, N.J. & Stuart, S.N. (1985). Threatened Birds of Africa and Related Islands: the ICBP/IUCN Red Data Book. International Council for Bird Preservation, Cambridge, U.K.
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. International Council for Bird Preservation, Cambridge, U.K.
Collar, N.J., Crosby, M.J. & Stattersfield, A.J. (1994). Birds to Watch 2: the World List of Threatened Birds. BirdLife Conservation Series 4, BirdLife International, Cambridge, U.K.
Collar, N.J., Wege, D.C. & Long, A.J. (1997). Patterns and causes of endangerment in the New World avifauna. Pp.237-260 in J.V. Remsen, ed. (1997). Studies in Neotropical Ornithology Honoring Ted Parker. American Ornithologists' Union (Orn. Monogr. 48), Washington, D.C.
Croxall, J.P. & Gales, R. (1998). An assessment of the conservation status of albatrosses. Pp.46-65 in G. Robertson & R. Gales, eds. (1998). Albatrosses: Biology and Conservation. Surrey Beatty & Sons, Chipping Norton, New South Wales.
Diamond, J.M. (1980). Patchy distributions of tropical birds. Pp.57-74 in M.E. Soulé & B.A. Wilcox, eds. (1980). Conservation Biology: an Evolutionary–Ecological Perspective. Sinauer Associates, Inc., Sunderland, Mass.
Diamond, J.M. (1984a). Historic extinctions: a Rosetta Stone for understanding prehistoric extinctions. Pp.824-862 in P.S. Martin & R.G. Klein, eds. (1984). Quaternary Extinctions: a Prehistoric Revolution. University of Arizona Press, Tucson.
Diamond, J.M. (1984b). “Normal” extinctions of isolated populations. Pp.191-246 in M.H. Nitecki, ed. Extinctions. Chicago: University of Chicago Press.
Dodson, C.H. & Gentry, A.H. (1991). Biological extinction in western Ecuador. Ann. Missouri Bot. Garden 78: 273-295.
Duckworth, J.W., Timmins, R.J. & Evans, T.D. (1998). The conservation status of the River Lapwing Vanellus duvaucelii in southern Laos. Biol. Conserv. 84: 215-222.
Erwin, T.L. (1988). The tropical forest canopy: the heart of biotic diversity. Pp.123-129 in E.O. Wilson, ed. (1988). Biodiversity. National Academy Press, Washington, D.C.
Frankel, O.H. & Soulé, M.E. (1981). Conservation and Evolution. Cambridge University Press, Cambridge, U.K.
Franklin, I.R. & Frankham, R. (1998a). How large must populations be to retain evolutionary potential? Anim. Conserv. 1: 69-70.
Franklin, I.R. & Frankham, R. (1998b). Response to Lynch and Lande. Anim. Conserv. 1: 73.
Howe, R.W., Davis, G.J. & Mosca, V. (1991). The demographic significance of “ sink” populations. Biol. Conserv. 57: 239-255.
Hutto, R.L. (1998). On the importance of stopover sites to migrating birds. Auk 115: 823-825.
ICBP (1992). Putting Biodiversity on the Map. International Council for Bird Preservation, Cambridge, U.K.
IUCN Species Survival Commission (1994). IUCN Red List Categories, as Approved by the 40th Meeting of the IUCN Council. IUCN–The World Conservation Union, Gland, Switzerland.
Johns, A.G. (1997). Timber Production and Biodiversity Conservation in Tropical Rain Forests. Cambridge University Press, Cambridge, U.K.
Kattan, G.H., Alvarez-López, H. & Giraldo, M. (1994). Forest fragmentation and bird extinctions: San Antonio eighty years later. Conserv. Biol. 8: 138-146.
King, W.B. (1978-1979). Red Data Book, 2: Aves. Second edition. International Union for Conservation of Nature and Natural Resources, Morges, Switzerland.
King, W.B. (1985). Island birds: will the future repeat the past? Pp.3-15 in P.J. Moors, ed. (1985). Conservation of Island Birds. Techn. Publ. 3, International Council for Bird Preservation, Cambridge, U.K.
Lambert, F.R. (1991). The conservation of fig-eating birds in Malaysia. Biol. Conserv. 58: 31-40.
Lambert, F.R. (1992). The consequences of selective logging for Bornean lowland forest birds. Phil. Trans. R. Soc. London B 335: 443-457.
Lowe-McConnell, R.H. (1987). Ecological Studies in Tropical Fish Communities. Cambridge University Press, Cambridge, U.K.
Lynch, M. & Lande, R. (1998). The critical effective size for a genetically secure population. Anim. Conserv. 1: 70-72.
Mack, A.L. & Wright, D.D. (1998). The Vulturine Parrot, Psittrichas fulgidus, a threatened New Guinea endemic: notes on its biology and conservation. Bird Conserv. Internatn. 8: 185-194.
de la Mare, W.K. (1987). On the definition of threats to the survival of species. Pp.113-121 in R. Fitter & M. Fitter, eds. (1987). The Road to Extinction: Problems of Categorizing the Status of Taxa Threatened with Extinction. International Union for Conservation of Nature and Natural Resources, Gland, Switzerland, and Cambridge, U.K.
Marsden, S.J. (1995). The Ecology and Conservation of the Parrots of Sumba, Buru and Seram, Indonesia. Ph.D. thesis, Department of Biological Sciences, Manchester Metropolitan University.
Marsden, S.J. (1998). Changes in bird abundance following selective logging on Seram, Indonesia. Conserv. Biol. 12: 605-611.
Mawson, P.R. & Long, J.L. (1994). Size and age parameters of nest trees used by four species of parrot and one species of cockatoo in south-west Australia. Emu 94: 149-155.
Milberg, P. & Tyrberg, T. (1993). Naïve birds and noble savages – a review of man-caused prehistoric extinctions of island birds. Ecography 16: 229-250.
Pimm, S.L., Moulton, M.P. & Justice, L.J. (1994). Bird extinctions in the central Pacific. Phil. Trans. R. Soc. London B 344: 27-33.
Potapov, E. (1992). Some breeding observations on the Siberian White Crane Grus leucogeranus in the Kolyma lowlands. Bird Conserv. Internatn. 2: 149-156.
Powell, G.V.N. & Bjork, R.D. (1994). Implications of altitudinal migration for conservation strategies to protect tropical biodiversity: a case study of the Resplendent Quetzal Pharomacrus mocinno at Monteverde, Costa Rica. Bird Conserv. Internatn. 4: 161-174.
Ralph, C.J. & van Riper, C. (1985). Historical and current factors affecting Hawaiian native birds. Bird Conservation 2: 7-42.
Robinson, S.K. (1998). Another threat posed by forest fragmentation: reduced food supply. Auk 115: 1-3.
Rothschild, W. (1907). Extinct Birds. Hutchinson & Co, London.
Savidge, J.A. (1987). Extinction of an island forest avifauna by an introduced snake. Ecology 68: 660-668.
Simberloff, D. (1986). The proximate causes of extinction. Pp.259-276 in D.M. Raup & D. Jablonski, eds. (1986). Patterns and Processes in the History of Life. Life Sciences Research Report 36, Springer-Verlag, Berlin.
Soulé, M.E. (1983). What do we really know about extinction? Pp.111-124 in C.M. Schonewald-Cox, S.M. Chambers, B. MacBryde & L. Thomas, eds. (1983). Genetics and Conservation: a Reference for Managing Wild Animal and Plant Populations. Benjamin/Cummings, Menlo Park, California.
Soulé, M.E., Bolger, D.T., Alberts, A.C., Wright, J., Sorice, M. & Hill, S. (1988). Reconstructed dynamics of rapid extinctions of chaparral-requiring birds in urban habitat islands. Conserv. Biol. 2: 75-92.
Stattersfield, A.J. (1997). Applying the new IUCN threatened species categories. Pp.67-70 in J. Baillie & B. Groombridge, eds. (1997). 1996 IUCN Red List of Threatened Animals. IUCN– The World Conservation Union, Gland, Switzerland.
Stattersfield, A.J., Crosby, M.J., Long, M.J. & Wege, D.C. (1998). Endemic Bird Areas of the World: Priorities for Biodiversity Conservation. Conservation Series 7, BirdLife International, Cambridge, U.K.
Steadman, D.W. (1995). Prehistoric extinctions of Pacific Island birds: biodiversity meets zooarchaeology. Science 267: 1123-1131.
Stresemann, E. & Grote, H. (1943). Ist Numenius tenuirostris im Aussterben begriffen? Orn. Monatsber. 51: 122-127.
Terborgh, J. (1974). Preservation of natural diversity: the problem of extinction prone species. BioScience 24: 715-722.
Terborgh, J. & Winter, B. (1980). Some causes of extinction. Pp.119-133 in M.E. Soulé & B.A. Wilcox, eds. (1980). Conservation Biology: an Evolutionary– Ecological Perspective. Sinauer Associates, Inc., Sunderland, Mass.
Thiollay, J.-M. (1992). Influence of selective logging on bird species diversity in a Guianan rain forest. Conserv. Biol. 6: 47-63.
Turner, I.M. (1996). Species loss in fragments of tropical rain forest: a review of the evidence. J. Appl. Ecol. 33: 200-209.
Wells, D.R. (1985). The forest avifauna of western Malesia and its conservation. Pp.213-232 in A.W. Diamond & T.E. Lovejoy, eds. (1985). Conservation of Tropical Forest Birds. Techn. Publ. 4, International Council for Bird Preservation, Cambridge, U.K.
Whitmore, T.C. (1997). Tropical forest disturbance, disappearance and species loss. Pp.3-12 in W.F. Laurance & R.O. Bierregaard, eds. (1997). Tropical Forest Remnants: Ecology, Management and Conservation of Fragmented Communities. University of Chicago Press, Chicago.