How Fast Are We Losing Avian Biodiversity?
Bird species are being driven Extinct by increasing human impacts on the planet. In total, 129 bird species have been classified as Extinct since 1500, with an additional four species listed as Extinct in the Wild, but surviving in captive populations (BirdLife International 2004b
). Additionally, 18 Critically Endangered species are considered Possibly Extinct by BirdLife International. (2004b
; see Materials and Methods
). Of these confirmed and likely extinctions, nine have occurred during the period 1988–2004 (BirdLife International, unpublished data). However, it is very difficult to produce robust estimates of recent extinction rates and to quantify how they have changed over short timescales. This is because extinction is difficult to detect once species become very rare (Diamond 1987
; Reed 1996
), and tiny populations of species potentially doomed by habitat loss or other threats may persist for many decades (Turner 1994
; Brooks et al. 1997
). For these reasons, extinction rates perform weakly as indicators of the current state of biodiversity (Balmford et al. 2003
By contrast, the RLIs presented here provide a robust, sensitive measure of the rate at which the world's birds are changing in relative projected extinction risk, as classified using the categories of the IUCN Red List. The indices show that the overall threat status of the world's birds has deteriorated steadily over the last 16 y. The RLI value has changed by −6.90% over this period. However, it should be noted that owing to the arbitrary nature of the weights applied to each category to calculate the score, this percentage decline is not directly comparable with percentage declines reported for population-based indices such as the Living Planet Index (Loh 2002
) or the United Kingdom headline indicator for wild bird populations (Gregory et al.2003
When the RLI is weighted by the relative extinction risk associated with each category in order to emphasise trends in the status of the most threatened species, the rate of decline of the index value appears to have levelled off in recent years (see ), owing to the number of such species deteriorating in status being balanced by the number improving. Nevertheless, it should be emphasised that one Critically Endangered species went Extinct in the Wild in the wild during the period (Hawaiian crow [C. hawaiiensis
]), and another is highly likely to have done so (Spix's macaw [C. spixii
]; BirdLife International 2004c
). These are potentially irretrievable losses to genetic diversity.
How can we interpret the RLI in relation to the CBD's target of reducing the rate of loss of biodiversity by 2010? The interpretation is different for measures of the state of biodiversity (e.g., total area of remaining forest) and measures of the rate of change in this state (e.g., annual percentage forest loss). For indices based on proportional change in a measure (plotted on a negative scale as with the RLI), if the measure is one of state, a significant diminution in downward trend would show that the target has been met. If the measure is one of rate of change of state, however, the target is not met until we see a positive trend, not just a decelerating decline. Some of the Red List criteria are based on absolute population size or range size, while others are based on rates of decline in these values or combinations of absolute size and rates of decline. These criteria are used to assign species to Red List categories that can be ranked according to relative projected extinction risk, and the RLI is calculated from changes between these categories. Hence an RLI value relates to the rate at which species are slipping towards extinction at a particular time. To show that the 2010 target has been met, the RLI must therefore show a positive trend. A downward trend, even if diminishing, shows that the slide of species towards extinction is accelerating, not slowing down. The negative trends in the RLI values (see ) thus show that in 2004 we are losing biodiversity at an increasing rate.
The RLIs show some interesting regional variations. The index for birds in the Indo-Malayan realm shows a sharp decline during the 1990s (see ). This was a result of the intensifying destruction of forests in the Sundaic lowlands of Indonesia, which escalated particularly in the late 1990s and led to predictions of almost total loss of lowland forest in Sumatra by 2005 and in Kalimantan by 2010 (Holmes 2000
; BirdLife International 2001
). As a consequence of these increasing rates of habitat loss, many species were uplisted to higher categories of threat under criterion A (rapid population declines). However, it is notable that there has been a significant deterioration in the threat status of birds of shrubland/grassland habitats as well as forest, and in the two other major ecosystems (freshwater and marine), indicating that birds in a broad spectrum of environments are under threat.
RLIs can be calculated for particular species groups that have specific conservation or policy significance. For example, there are particularly active conservation networks for game birds (e.g., World Pheasant Association), raptors (e.g., World Working Group on Birds of Prey), and parrots (e.g., Loro Parque and World Parrot Trust), and the threat status of all three of these species groups is deteriorating, with steeper declines in the index value for parrots in the earlier part of the period (see ). In addition, there are several international conservation treaties targeting particular suites of species (the Ramsar Convention on Wetlands, the Convention on Migratory Species [CMS], and the Agreement on the Conservation of Albatrosses and Petrels [ACAP] under the CMS) for which disaggregated RLIs provide a metric against which to judge their success in improving the fortunes of the species involved. The RLI for albatrosses and large petrels shows how dramatically their threat status has deteriorated in recent years (see ). This is closely linked to the expansion of commercial longline fisheries (both legal and illegal), which causes incidental mortality of albatrosses and other seabirds when they get caught on baited hooks and drown (Tuck et al. 2001
; BirdLife International 2004b
). The total reported effort from fleets in the southern oceans has been well over 250 million hooks per year since the early 1990s, with some fleets expanding rapidly in the last decade (Tuck et al. 2003
). Models for at least some albatross species show clear links between population declines and these increases in longline fishing effort (Tuck et al. 2001
). Mitigation measures are effective (Løkkeborg 2001
), and the RLI will provide a useful measure by which to judge the effectiveness of the implementation of ACAP, following its entry into force in 2004.
It should be noted that setting all disaggregated index values to a common baseline in 1988 obscures any changes prior to this period (see, e.g., Pauly 1995
). For example, although the Indo-Malayan realm has shown the most severe recent index declines, ‘only' six extinctions occurred there between 1500 and 1988, whereas at least 62 bird species are known to have gone Extinct in the Australasian/Oceanic realm during the same period, and 40 in the Afrotropical realm (BirdLife International 2000
). Similarly, the terrestrial ecosystem has suffered far more extinctions since 1500 (115) than the freshwater (17) or marine ecosystems (five), but all are set to a common baseline in 1988.
The RLI is based on the number of species in each Red List category. In order to make the index sensitive, not just to the total number of threatened species, but also to the changes in category assigned to each species, each category was given a weighting. We used an ‘equal-steps' approach (with incremental increases from one for Near Threatened through to five for Extinct) to reflect the ordinal ranks of the categories, whereby each step from Least Concern to Extinct indicates that at least one measure of extinction risk has become worse. The advantage of this approach is that it is simple, and the trends in the resulting index are driven by a relatively large number of species (hence producing a more robust and representative index). This is because a species moving from Least Concern to Near Threatened contributes just as much to the changing score as a Critically Endangered species going Extinct, and the numbers of species in each category (and the number moving in and out of each category) increases disproportionately from Critically Endangered to Least Concern (see ).
However, steps between lower categories of threat represent smaller increases in extinction risk than steps between higher categories. Therefore we also tested an ‘extinction risk' approach, with each category weighted according to its relative extinction risk based on the quantitative thresholds for each of its criteria (). Although this approach also relies on some assumptions (e.g., about the type of extinction risk curve, and the extinction risk associated with Near Threatened), it is based on the principles of extinction dynamics, in contrast to the equal-steps approach.
Weights for Red List Categories Critically Endangered, Endangered, and Vulnerable, Based on Relative Extinction Risk Associated with Various Red List Criteria
The most important difference between the two approaches is the effect of status changes in less-threatened or nonthreatened species. The equal-steps approach gives an index that is heavily influenced by movements of species among the lower categories of threat. The extinction risk approach gives an index that is largely influenced by movements of species among the higher threat categories. For example, if a Vulnerable species improves in status and becomes Near Threatened, and at the same time, a Critically Endangered species goes Extinct, the RLI based on equal-steps weights registers no change, but the index based on extinction risk weights shows a substantial decrease. Downlisting of a Vulnerable species to Near Threatened might represent a very substantial population increase, whereas extinction of a Critically Endangered species might represent the loss of very few individuals. The latter is arguably more significant in terms of genetic diversity, but the former might be more important as an indicator of wider biodiversity trends. Thus, the extinction risk weights emphasise the loss of biodiversity owing to imminent or potential extinctions of species, whereas the equal-steps weights allow the index to capture large changes in the populations of less-threatened species.
For the RLI for complete taxonomic groups, and for disaggregating the index to show trends for subsets of species, for example, in particular realms or ecosystems we used the equal-steps approach because the number of species moving between the higher threat categories (those effectively driving trends when an ‘extinction risk' weighting is used) was too small to be meaningful in disaggregated indices. Only 23% of all genuine status changes (58 species in total) involved moves in or out of the highest threat categories. However, for examining trends in the species closest to extinction, we used the weights based on relative extinction risk.
Weaknesses of RLIs
The usefulness of the IUCN Red List as an indicator of trends in the status of biodiversity (e.g., Smith et al. 1993
) has been previously questioned on the grounds that (1) the categories are subjective; (2) taxonomic treatment is uneven, and listings are biased towards attractive, spectacular, high-profile, or better-known species; and (3) most species move between categories because of changes in knowledge or taxonomy, not as a consequence of genuine improvement or deterioration in status (Cuarón 1993
; Burgman 2002
; Possingham et al. 2002
; but see Lamoreux et al. 2003
The first of these problems has been addressed since 1994, when quantitative and objective categories and criteria for the IUCN Red List were introduced (IUCN 1994
). The second problem can be overcome by calculating indices only for taxonomic groups in which all species have been comprehensively assessed and reassessed (as shown here) or by developing indices based on a stratified sample from diverse taxonomic groups (see below). The third problem has already been addressed because since 2001 the IUCN Red List has required clear documentation of the reason for any reclassification (IUCN 2001
). Hence, movements of species between categories owing to knowledge, taxonomy, or other ‘nongenuine' reasons can be easily excluded when calculating the index.
RLIs have a fairly coarse level of resolution of status changes because of the broad nature of Red List categories. Populations in the wild may have to undergo quite significant changes in size, trend, or distribution before crossing the thresholds to qualify for a higher or lower Red List category and, hence, before changing the RLI value. This is inherent in using the Red List categories rather than more precise parameters such as estimates of population size. It is not always true, however: The Red List criteria allow for species to be assessed as threatened on the basis of projected declines, and thus changes in status can reflect new or emerging threats in anticipation of population or range changes. We suggest that the disadvantage of coarse resolution is outweighed by the advantage of using a system that allows all the world's species in a taxonomic group to be assessed, rather than just a (potentially biased) subset for which adequately detailed information is available.
Insensitivity of the index to status changes may also arise from time lags between changes in a species' population or range and changes in the RLI value, because of delays before detection of the status change, and/or before this knowledge becomes available to assessors. This is potentially more problematic, but several factors act to mitigate it. The Red List Programme partners have a large and expanding network of scientists across the world providing detailed and up-to-date information for an increasing number of species. Furthermore, with improving channels of communication (in particular, the increasing use of the World Wide Web to solicit information, for example, BirdLife's Web-based Globally Threatened Bird discussion forums; BirdLife International 2004a
), we expect that such delays will diminish. For birds, the data support this: whereas just 42% of genuine status changes between 1988 and 1994 were detected in 1994 (with 43% detected during 1994–2000 and 15% detected during 2000–2004), 88% of changes during 1994–2000 were detected in 2000, and just 12% were detected in the subsequent 4 y. Using the data from the 1994–2000 period (because information gathering has improved considerably since 1988–1994), we can estimate the likely number of genuine status changes for 2000–2004 that have not yet been detected (six; see Materials and Methods
) and, hence, estimate the possible degree of error associated with the 2004 RLI value. The results show that it may be an under- or overestimate by 0.21%–0.37% (see ): a small and acceptable margin of error. In future, we anticipate smaller retrospective adjustments to the index values, and a smaller and more predictable error associated with the most recent index value. The major advantage of backtracking status changes to the appropriate time period is that the index trends do not get distorted by the belated discovery of genuine status changes, which, for example, might result from the exhaustive research that takes place when a Red Data Book is published (e.g., BirdLife International 2001
). This arguably outweighs the disadvantage that the slope of the index between two particular dates may change slightly in future releases of the index.
How robust are RLIs? A potential criticism is that they are based on status changes in small numbers of species. However, between 1988 and 2004 the RLI declined by a degree equivalent to almost 10% of species in the categories Near Threatened to Critically Endangered deteriorating in status sufficiently to be uplisted by one category to a higher category of threat. Although relatively few in number (250), these status changes are the most important among the world's birds in terms of changes in projected extinction risk. We therefore suggest that the declines shown by the RLI since 1988 represent very significant losses to global biodiversity.
Relatively large numbers of species changed categories in 1994 and 2000 owing to improvements in knowledge and improved consistency of interpretation of information against the Red List criteria (see ). This was because of the introduction of quantitative criteria for assigning species to categories in 1994 (Collar et al. 1994
; IUCN 1994
) and the mapping of all threatened species and more rigorous justification for Near Threatened status in 2000 (BirdLife International 2000
). By 2000–2004, only 6.7% of threatened and Near Threatened species changed category owing to improved knowledge (see ). Nevertheless, it is true that a small proportion of species may be sufficiently poorly known that there is uncertainty over their status and whether this has changed over time. If this introduces any bias, it may be towards an overoptimistic RLI trend. This is because well-studied species (with better data and hence more certain Red List assessments) may be more likely to be those receiving conservation attention and, hence, improving in status (or at least deteriorating less rapidly). All data used in Red List assessments for birds (e.g., population size, trends, etc.) are coded for data quality, and in future the RLI will also be calculated separately for species with high-quality data, in order to test whether such biases exist.
Strengths of RLIs
The greatest strength of the RLIs presented here is that they are based on comprehensive and complete assessments of nearly all species in a taxonomic group across the world (just 0.8% of birds are listed as Data Deficient and hence excluded from the calculation of the RLI). Most other global indicators based on, for example, population estimates, are derived from data biased towards common, well-studied species in the developed world, particularly Europe and North America. For example, the Living Planet Index (Loh 2002
) is based on indices for populations in marine, freshwater, and forest ecosystems. However, 70% of the 195 populations contributing to the freshwater ecosystem are in Europe or North America, while just 18% of the 282 populations contributing to the forest ecosystem index are in the tropics, where the greatest biodiversity is found (Loh 2002
). Similarly, in a global index based on data from 936 amphibian populations from 37 countries around the world, 89% of populations (835) were from Europe or North America, and just 2.2% (21) were from Asia and 5.5% (51) from South/Central America (Houlahan et al. 2000
). By contrast, the RLI for birds is based on trends for nearly all the world's 10,000 bird species. RLIs for other completely assessed taxonomic groups are in development (see below).
At present, indicators based on more representative suites of species are only available for particular countries or regions, such as the United Kingdom headline indicator for birds (Gregory et al. 2003
) and the Pan-European Common Bird Index (BirdLife International 2004b
; Gregory et al. 2004
). Indices based on population trends (particularly at the regional scale) generally include few species that are rare, localised, or difficult to survey, including those most susceptible to extinction. RLIs can incorporate status changes in such species because the Red List process is an effective system for making meaningful inferences from data that are imprecise or incomplete.
Species-based indicators such as the RLI arguably provide far more powerful measures of biodiversity loss than other indicators proposed for measuring progress towards the 2010 target (CBD 2004
). Trends in the extent of biomes and habitats are of necessarily coarse resolution and take no account of the distribution of biodiversity within and between habitats; trends in the genetic diversity of domesticated animals and cultivated plants provide measures related to only a tiny proportion of biodiversity; and trends in the coverage of protected areas are a measure of responses to biodiversity loss rather than a measure of the state of biodiversity.
At present, data are only available for birds to produce the sorts of indices shown here. By 2010 at least two complete global assessments will also be available, and RLIs calculated for all the world's mammals (about 5,000 species), amphibians (about 5,700 species), and hopefully some plant and marine groups. Additional indices, and an aggregation of RLI trends in multiple groups, will provide a more representative picture of the changing state of biodiversity. In recognition that this will take some time to implement, the IUCN Red List Programme is also developing a sampled RLI based on a stratified sample of about 3,000 species from all major taxonomic groups, biogeographic realms, ecosystems, and Red List categories. This will provide an index that may be more representative of trends in the threat status of all biodiversity. We suggest that RLIs will have a key role to play alongside other types of indicators in assessing progress towards reducing the rate of, or halting, the loss of biodiversity.