Photo 5.1
The Keeled Box Turtle Pyxidea mouhotii (Endangered) inhabits the forest floor leaf litter of localized areas of evergreen forests from northeastern India through Myanmar, Lao PDR and Viet Nam to southern China. The species has been harvested in large numbers for the Asian turtle consumption trade, as well as for the international pet trade. Populations have disappeared and remaining populations, including those in formally designated protected areas, appear to be declining.
Photo: © Peter Paul van Dijk.
The geographic range of a species can be assessed using a variety of techniques (Gaston 1994). First, and at the coarsest resolution, species distributions have traditionally been mapped through known occurrence in predefined geographic units, such as countries (Mittermeier 1988) or geopolitical units (Brummitt 2001), and ecological systems and biomes (Olson et al. 2001). Second, are polygon range maps (“Extent of Occurrence” (EOO) defined in IUCN (2001) and Appendix 2e), based on a combination of known records and specialist knowledge, although these greatly overestimate occupancy within the range (Corsi et al. 2000). Third, the finest resolution approach is to compile point data– known point occurrences (often supported by museum or herbarium specimens) of a given species in a given place at a given time – but uneven sampling is a serious problem for the assessment of point data (Peterson et al. 1998; Peterson and Watson 1998). These sampling problems can be reduced by development of inductive range models (Peterson et al. 2002) or the establishment of grid based sampling systems to produce atlas data (Udvardy 1981). Data limitations mean that it has not yet been possible to use these latter two approaches across entire taxonomic groups, worldwide.
This section presents a geographic analysis of the IUCN Red List data through these three approaches. The first –counting species occurrences in predetermined geographic units (Section 5.2) – has been used in previous analyses of the Red List (Baillie and Groombridge 1996; Hilton-Taylor 2000). The second – analysis of EOO range maps (Section 5.3) for individual species – has previously been possible only for threatened birds (BirdLife International 2000), and so the incorporation of EOO data in the documentation required when submitting a species assessment for inclusion on the IUCN Red List allows us remarkable new insights. The third – mapping known point occurrences – is still in its infancy, despite its importance to conservation on the ground (Collar 1993–94, 1996), but here we are able to illustrate the potential provided by mapping selected species at the site scale.
The distribution of threatened species is summarized here following five predetermined geographic classifications. One of these is political (countries), the other four ecological (ecological systems, biogeographic realms, biomes, and habitats). The occurrence of threatened species in all of these (with the exception of biomes) is evaluated by using the relevant required documentation for including species on the IUCN Red List (see IUCN 2001, pp. 27–29).
The richness of threatened species per country is useful to give context as a coarse measure of threats to biodiversity, but is heavily conflated by area (Balmford and Long 1995) and driven by the occurrence of widespread species (Lennon et al. 2004). Dividing threatened species richness by total numbers of species per country does give a useful measure of relative threat to a nation's biodiversity. The presentation of threatened species occurrences by country is also useful in providing a crosscheck to national Red Lists, and vice versa (Hilton-Taylor et al. 2000; Rodriguez et al. 2000), given the important implications of these for national conservation policy (see Section 8). Particularly informative are the numbers of threatened species endemic (see Appendix 2e for a definition of this term) to each country, because they can guide a “doctrine of ultimate responsibility” for each nation's contribution to global biological heritage (Mittermeier et al. 1998).
In Appendix 3j, using the country occurrence documentation data from the 2004 IUCN Red List, we list the numbers of threatened species and threatened national endemic species per country for the six taxonomic groups for which the coverage of species in the Red List is most complete: mammals, birds, amphibians, turtles, conifers and cycads; a seventh group, the chondrichthyan fishes (sharks, rays and chimaeras), are also included because although only a third of species have been assessed, it is the largest marine group currently on the Red List. Comparative geographic analyses are impossible for groups that have not been comprehensively assessed. For example, endemic vascular plants have been comprehensively assessed in Ecuador (Valencia et al. 2000) and a handful of other countries, but not globally, and so any analysis of threatened plant distributions would be greatly biased towards such countries. We, therefore, restrict data assessment and presentation to those taxonomic groups within which data from all species can be analysed.
Photo 5.2
Siphocampylusecuadoriensis (Endangered) is ashrub endemic to theEcuadorian Andes. The species isthreatened by theongoing conversionof native vegetationto pasture andongoing deforestationwithin protected areaswhere it occurs.
Photo: © Suzanna León-Yánez.
Several overall patterns stand out from Appendix 3j. First, there is a rather high level of correspondence between the numbers and, especially, proportions of threatened species per country, for each of the seven taxonomic groups. The main exceptions to this pattern appear to be driven by ecological constraints. For example, amphibians have very poor dispersal abilities over saltwater, and so do not occur naturally in many oceanic island nations important for threatened birds such as Mauritius or Vanuatu. Second, while those countries with the largest numbers of threatened and threatened endemic species lie in the continental tropics, those with the highest proportions of threatened endemics are generally tropical island nations (such as Cuba, with >50% of threatened species endemic for five of the seven taxa considered here). This is a combined result of the low species richness of islands and the ecological naïveté of those species that do occur on islands (Diamond 1991). The exception to this pattern is amphibians, for which threatened species have such tiny ranges that nearly all are national endemics.
Considering Appendix 3j country-by-country, the exceptional importance of five countries, Australia, Brazil, China, Indonesia, and Mexico, stands out. Other countries or territories holding particularly large numbers of threatened species include Colombia, India, New Caledonia, Peru, South Africa, and Viet Nam (all of these are among the top three countries for at least one taxonomic group) while Colombia, India, Malaysia, Myanmar, New Caledonia, Papua New Guinea, the Philippines, South Africa, and the United States are all among the top three countries for numbers of threatened endemics for at least one taxonomic group. Additional countries characterized by particularly high proportionate threat in multiple taxa include Madagascar, São Tomé and Principe, and the Seychelles.
Photo 5.3
Although Brazil currently has 22% of its 95 primate specieslisted as threatened, one of these, the Golden Lion TamarinLeontopithecus rosalia (Endangered), is showing signs ofrecovery after nearly 30 years of conservation efforts.
Photo: © Juan Pratginestós/WWF-Brazil.
Photo 5.4
In Sri Lanka, the loss of forest cover has impacted manyspecies including the Slender Loris Loris tardigradus (Endangered).
Photo: © Anna Nekaris.
The most straightforward framework for assessing the ecological distributions of threatened species is to divide the planet's surface into three systems: terrestrial, freshwater, and marine. This classification is complicated by those species that live in the interface between systems (e.g., shorebirds) and those that live in multiple systems (e.g., many amphibians). These are a small proportion of species overall, however – most species occur only in one of the three ecological systems.
The numbers and proportions of total and threatened mammals, birds, amphibians, turtles, chondrichthyan fishes (sharks, rays and chimaeras), conifers and cycads occurring in each of the three ecological systems (as coded on the IUCN Red List) are shown in Table 5.1. The absolute number of threatened species known from marine systems is low, primarily a reflection of the recording biases towards terrestrial and freshwater taxa. Even including those taxonomic groups for which comprehensive assessments have yet to be conducted yields only 187 threatened marine species, as opposed to 4,427 that live on land. Assessments in freshwater systems are rather further ahead than marine systems: 1,388 freshwater species (including taxa not yet comprehensively assessed) are listed as threatened. This is wholly an artefact of the fact that major aquatic groups (above all, fish) have yet to be comprehensively assessed for the Red List.
Table 5.1
Total numbers of species and numbers ofthreatened species from well-assessedtaxonomic groups occurring in marine, freshwater and terrestrial ecological systems (with proportions indicated in parentheses). Some species occur in more than one system, and so are counted multiple times in the table.
Considering the proportions of threatened species per system gives a rather different picture, without any clear patterns. Among mammals, a considerably higher proportion of freshwater species is threatened than of marine or terrestrial species. Among birds, turtles and chondrichthyan fishes, on the other hand, a considerably higher proportion of marine species is threatened than of freshwater or terrestrial species. A slightly higher proportion of terrestrial amphibians are threatened than of freshwater species. Across all seven taxa, the proportion of threatened freshwater species is slightly higher (25%) than that for marine and terrestrial systems (22 and 21% respectively), as expected from the intensity of threat to freshwater (McAllister et al. 1997).
Biogeographic realms are the eight continent-scale terrestrial and freshwater regions distinguished by characteristic biota that reflect shared evolutionary histories (Udvardy 1975). In Table 5.2 the numbers and proportions of threatened mammals, birds, and amphibians occurring in and endemic to each of the eight biogeographic realms are summarized.
Much the greatest numbers of threatened species, for all taxa, occur in the tropical continents: the Neotropical, Afrotropical, and Indomalayan realms. The Australasia and Palearctic realms have many less threatened species, the Nearctic less still (although it has more threatened amphibians than Australasia), and the Antarctic realm has almost no threatened species. Oceania, while having a low richness of threatened species, has a remarkably high proportionate threat, reflecting again the vulnerability of oceanic island biodiversity. Proportionate threat is remarkably similar between biogeographic realms and taxa, although rather low for Nearctic mammals and Australasian amphibians, and high for Indomalayan and Neotropical amphibians.
Table 5. 2
Total numbers of species, and numbers and percentages of threatened species, from well-mapped taxonomic groups (mammals, birds, amphibians) occurring in and (in parentheses) endemic to each of the eight terrestrial biogeographic realms. Marine mammals are excluded.
The Food and Agricultural Organization (FAO) of theUnited Nations, which has defined 19 marine regions (Fishing Areas) worldwide, provide a marine analogue of thebiogeographic realms. Coding against these marine regionsis required documentation when submitting assessments ofmarine species for inclusion on the IUCN Red List (IUCN2001), and so for those few marine megafaunal groups thathave been comprehensively assessed, it is therefore nowpossible to begin to unveil geographic patterns in thegeography of threat in the marine system. Table 5.3 summarizes the numbers of threatened marine mammals, seabirds, chondrichthyan fishes, and seahorses in each FAOFishing Area (we cannot assess relative threat because thelatter two groups have not yet been comprehensivelyassessed, but sufficient species have been assessed forcomparative purposes).
Relative to the other three groups, marine mammalsappear to be an outlier with much the largest numbers ofthreatened species occurring in the northern PacificOcean regions. By contrast, the regions holdingthe greatest numbers of threatened seabirds, chondrichthyan fishes and seahorses are concentrated inthe ‘coral triangle’ region of the eastern Indian Oceanand southwest and western central Pacific. The Arcticand Antarctic Oceans hold the fewest threatened speciesacross all taxa (with the exception of threatenedseabirds, a number of species of which occur inAntarctica). The results for the chondrichthyan fishesneed to be treated as preliminary as the concentrationsof threatened species may simply be a reflection ofwhere assessment workshops have been held to date (see Box 2.3).
Table 5. 3
Numbers of threatened marine mammals (cetaceans, seals and sirenians), seabirds, chondrichthyan fishes (sharks, rays and chimaeras), and seahorses in each FAO Fishing Area. Note: only marine mammals and seabirds have been comprehensively assessed.
Photo 5.5
The Northwest Pacific (Asia) Grey Whale Stock Eschrichtius robustus (Critically Endangered) is geographically distinct, and is thought to have less than 50 reproductive individuals. This subpopulation was hunted to near extinction and remains severely depleted.
Photo: © David W. Weller.
Photo 5.6
TheWandering Albatross Diomedea exulans (Vulnerable) is a wide-ranging pelagic species of the southern oceans, but most of its breeding colonies are on Subantarctic islands. The main cause of decline is the impact of longline fisheries.
Photo: © Tony Palliser / BirdLife International.
At a finer scale, it is possible to assess the distributions of threatened species across biomes. Biomes represent globalscale variation in the structure, dynamics and complexity of terrestrial and freshwater communities and ecosystems that are driven by key global-scale patterns such as temperature and precipitation. Olson et al. (2001) identified 14 biomes worldwide, a classification followed here in the interest of standardization. In Figures 5.1 and 5.2 we graph the numbers and proportions of threatened mammals, birds and amphibians occurring in and endemic to each biome.
Figure 5.1
Numbers of threatened mammals, birds and amphibians occurring in each biome (proportion of threatened species indicated in red).
Figure 5.2
Numbers of threatened mammals, birds and amphibians endemic to each biome (proportion of threatened species indicated in red).
The results of this analysis are stark: Tropical/ Subtropical Moist Broadleaf Forest is far and away the richest biome in terms of numbers of species and of threatened species for all three taxa, and is to a first approximation the only biome holding significant numbers of endemics or of threatened endemics for any of the three taxa. Tropical/Subtropical Dry Broadleaf Forest, Tropical/Subtropical Grassland, Savanna and Shrubland, Montane Grassland and Shrubland, and Desert and Xeric Shrubland all hold moderately large numbers of species and of threatened species for all three taxa (with the exception of the latter, which holds only a handful of threatened amphibians). The high-latitude biomes of Boreal Forests/Taiga and Tundra hold very few species, and even the Mediterranean Forest, Woodland and Scrub are remarkably depauperate (though this will probably not be the case once comprehensive plant data are included).
Photo 5.7
The Lesser Florican Sypheotides indica (Endangered) is a dry grassland species from India, Nepal and Pakistan. It has a very small, declining population, primarily a result of loss and degradation of its habitat.
Photo: © Asad Rahmani.
The finest ecological scale at which one can assess the distribution of threatened species is the scale of habitats, and, indeed, coding species up to their habitat preferences is part of the required documentation in the Red List assessment process (IUCN 2001). The importance of each major habitat for threatened and non-threatened birds and amphibians is illustrated in Figure 5.3 (these data have yet to be comprehensively compiled for any other taxa). Not surprisingly, given the results above for biomes, forest habitats are clearly the most important across both taxa. Grassland and shrubland habitats also hold high numbers of species. For amphibians, inland wetland habitats are exceptionally important, particularly for those species that have a larval stage. Another interesting pattern to emerge is the tendency of both bird and amphibian species to use artificial habitats (both terrestrial and aquatic), although analyses of the relative importance of each habitat type for birds reveals that these are nonetheless of minor importance (BirdLife International 2004b). Not surprisingly, considering the bias here towards terrestrial vertebrates, marine habitats come out as having few species (with the exception of marine representatives of the birds), as do desert habitats (which would likely come out stronger were data on reptiles and plant groups such as cacti included). When one considers the proportion of threatened species in a habitat as a percentage of the total occurring, marine emerges as particularly important for birds, almost certainly a result of the threatened status of, in particular, species in the order Procellariiformes (albatrosses, petrels and shearwaters). For amphibians, forest habitats clearly emerge as holding the largest numbers of threatened species, although at least 28% of amphibians in freshwater habitats are considered threatened.
Figure 5.3 The importance of each major habitat for birds (9,407 species) and amphibians (5,708 species), showing the number of threatened species (in red) relative to the totals. Extinct species are excluded.
Analysis of Extent of Occurrence (EOO) data has never before covered more than a single continent, for example, Africa (mammals, birds, snakes and amphibians; Brooks et al. 2001). However, such data are now available worldwide for all mammal, turtle and amphibian species (as part of the documentation required for assessing the status of species for the IUCN Red List). Data for non-threatened, Old World birds are not yet available, but all threatened (BirdLife International 2000, updated 2004a) and Western Hemisphere (Ridgely et al. 2003) species now have EOO data compiled.
In Figure 5.4, the species richness of all species of mammals, Western Hemisphere birds, freshwater turtles, and amphibians is mapped. Species richness patterns are primarily driven by the distributions of common, widespread species (Lennon et al. 2004), but they nonetheless provide context for threatened species distributions (Section 5.3.3).
The most obvious pattern from these data is that for all taxa the tropics hold much higher species richness than do the temperate, boreal and polar regions. Figure 5.5 demonstrates this by plotting the number of species in each 5-degree latitudinal band for all mammals, threatened birds, and amphibians. As expected from the relationship between the number of species in an area and the size of that area (Rosenzweig 1992), some of this pattern is explained by variation in landmass across latitudinal bands. However, species richness is much higher in the tropics than would be expected based on area alone, peaking around the equator for all taxa.
The other pattern obvious from Figure 5.4 is one of covariance between taxa. Thus, for example, species richness per grid cell is tightly correlated between mammals, freshwater turtles, and amphibians. Considering the Western Hemisphere only yields similarly high correlation coefficients for each of these three taxa compared to birds. Obviously, there are taxon specific differences driven by particular biological traits. Birds, for example, have the ability to disperse over water more than most of the taxa mapped here, and so occur in larger numbers on islands, while ectothermic (cold-blooded) reptiles flourish in desert regions generally depauperate in other animal taxa. Other differences are less easily explained, such as the high richness of mammal species in East Africa, and of turtles and amphibians in the southeastern USA. In general, these differences will increase with increasing evolutionary (and hence often corresponding ecological) difference between taxa (Reid 1998); one expects less correlation between mammal and plant distributions, for instance, than one would between mammal and bird distributions.
Figure 5.4 Species richness maps for four taxonomic groups: clockwise from top left: mammals (marine species excluded); Western Hemisphere birds; amphibians; and turtles (at a half-degree resolution). Dark red colours correspond to higher richness, dark blue to lowest. Colour scale based on 20 equal-area classes. Maximum richness equals 258 for mammals, 877 species for Western Hemisphere birds, 21 species for turtles, and 142 amphibians.
Figure 5.5 Number of species in each 5-degree latitudinal band for all mammals, threatened birds, and all amphibians.
There is a widespread correlation between species' range size and extinction risk (Purvis et al. 2000b), and, indeed, geographic range is inherent in the Red List Criteria (see Appendix 2a). It is well known that the frequency distribution of species' range sizes has a strong right skew; most species have small range sizes (Gaston 1996). Absolute values vary within this general pattern. More mobile species, such as birds, tend to have large distributions, while those of very sedentary species, such as amphibians, are generally much smaller (Figure 5.6). Nevertheless, the shape of frequency distributions of species' range sizes appears to be similar across all taxa examined to date (with the median range size consistently an order of magnitude smaller than the mean), probably because consistent processes are shaping these distributions (Gaston 1998).
Not only do most species have small ranges, but also these narrowly distributed species tend to co-occur in ‘centres of endemism’ (Anderson 1994). In Figure 5.7 a threshold approach is applied to map the centres of endemism inhabited, respectively, by more than two overlapping mammal, bird and amphibian species (freshwater turtles are also included, for comparison) with global distributions of less than 50,000 sq. km. The most consistent pattern emerging is that almost all centres of endemism lie in isolated or topographically varied regions. This is true for both geographical isolates such as mountains, peninsulas and islands. Perhaps as a consequence of this, they also tend to be near the coast. Another consistent pattern revealed by Figure 5.7 is the extreme concentration of centres of endemism in the tropics. This is the geographical manifestation of ‘Rapoport's rule’ (Rapoport 1982), which states that the mean latitude of a species' range correlates with the species' range size, although the generality of this ‘rule’ has been questioned (Gaston 1999) and it may be explicable by chance alone rather than by any underlying biological cause (Colwell and Hurtt 1994). The degree of overlap between centres of endemism across birds, mammals and amphibians is remarkable (Figure 5.8).
Figure 5.6
Frequency distribution of log10 transformed range sizes for mammals (marine species excluded; 4,734 species, mean = 1. 7 x 106 km2; median = 2. 5 x 105 km2), birds (species endemic to the Western Hemisphere only; 3,980 species, mean = 2. 1 x 106 km2; median = 4. 0 x 105 km2), and amphibians (4,409 species, mean = 4. 0 x 105 km2; median = 1. 8 x 104 km2). Data Deficient species are excluded. The log10 transformation makes the distribution look slightly left-skewed, but in fact the untransformed distribution is strongly right-skewed, that is, most species have very small range sizes.
Just as species are not evenly distributed across the planet, so threats to species are not evenly distributed (Sanderson et al. 2002). The species richness of threatened mammals, birds, freshwater turtles, and amphibians is illustrated in Figure 5.9. The maps show interesting similarities and differences between the groups. All four taxa show marked concentrations of threatened species in southern Brazil, Madagascar, the Western Ghats of India, the eastern Himalayas, central China, mainland Southeast Asia, Sumatra, Borneo, and the Philippines; threatened mammals, birds and amphibians are also concentrated in the Andes, West Africa, Cameroon, the Albertine Rift of Central Africa, the Eastern Arc Mountains of Tanzania, and Sri Lanka. Somewhat not surprisingly then, these same regions have all been identified as “biodiversity hotspots” (Myers et al. 2000; Mittermeier et al. 2004). All of these patterns are heavily driven by the maps of restricted-range species (Figure 5.7).
Figure 5.7 Centres of endemism inhabited, respectively, by more than two overlapping species with global distributions of less than 50,000 sq. km (mapped at a quarter-degree cell). Clockwise from top left: mammals; birds; amphibians; and turtles.
Figure 5.8 Overlap of centres of endemism inhabited, respectively, by more than two overlapping mammal, bird, and amphibian species with global distributions of less than 50,000 sq. km (mapped at a quarter-degree cell). This map excludes freshwater turtles (shown above), which have not yet been comprehensively assessed.
Photo 5.8
Hyperolius rubrovermiculatus (Endangered), one of the African reed frogs, is only known from the Shimba Hills in coastal Kenya, where it is intrinsically at risk because of its small range.
Photo: © de Saix.
Photo 5.9
Although the Grey-necked Picathartes Picathartes oreas (Vulnerable) has a relatively wide range, its population throughout west-central Africa is highly fragmented, and considered small and possibly in overall decline.
Photo: © Tasso Leventis.
Photo 5.10
Pygmy Hog Sus salvanius (Critically Endangered) found in the tall grasslands of the northern Indian subcontinent. Numbers remaining in the wild are very low and all hopes rest on a successful breeding and reintroduction programme. This juvenile animal was photographed at the Pygmy Hog Research and Breeding Centre in Basistha (near Guwahati, Assam, India).
Photo: © Roland Seitre.
Figure 5.9 Threatened species richness maps for four taxonomic groups clockwise from top left: mammals (marine species excluded); birds; amphibians; and turtles (at a half-degree resolution). Dark red colours correspond to higher richness, dark blue to lowest. Colour scale based on 10 equal-area classes. Maximum richness equals 25 species for mammals, 25 species for birds, 16 species for turtles, and 44 species for amphibians.
The mammal map (Figure 5.9) is noteworthy in that there is at least one threatened mammal species in most parts of the world. This is probably a reflection of the propensity for large-bodied, widely distributed mammals to become globally threatened. In addition to the geographic regions listed above, important concentrations of threatened mammals also occur in the eastern Amazon basin, southern Europe, Kenya, Sumatra, Java, the Philippines, New Guinea and Australia. Interestingly, Mesoamerica and the Caribbean islands are relatively unimportant for threatened mammals (in the case of the Caribbean, this is probably due to past extinctions (see Section 3)), but, on the other hand, clearly stand out for amphibians.
The bird map differs from the others in that the importance of oceanic islands is emphasized. Other areas that are of great importance for threatened birds, but which are not listed above, include the Caribbean islands, the Cerrado woodlands of Brazil, the highlands of South Africa, the plains of northern India and Pakistan, Sumatra, the Philippines, the steppes of central Asia, eastern Russia, Japan, southeastern China, and New Zealand. As for mammals, Mesoamerica and Australia are relatively less important. However, in contrast to mammals, the Amazon basin, Europe, Java and New Guinea are of relatively lower importance for threatened birds.
Threatened freshwater turtles exhibit rather different species richness patterns than the other taxa. In addition to the fact that their species richness is very low in the Atlantic Forest, Cerrado, Tropical Andes, Guinean Forests of West Africa, Eastern Arc Mountains and Coastal Forests and other hotspots holding so many threatened mammals, birds, and amphibians, they also concentrate in some surprising areas. These include the Amazon (due to the presence of the large, wide ranging, and heavily persecuted river turtles of the genus Podocnemis), the eastern and southwestern United States, and Asia Minor.
The most noteworthy aspect of the amphibian map, is that most of the world is devoid of threatened amphibian species (the opposite situation to that of mammals). However, threatened amphibians occur more densely in smaller areas than either mammals or birds (up to 44 species per half degree grid square, compared with 24 for both mammals and birds). The fact that concentrations of threatened amphibians are often in tiny areas of the montane tropics makes it hard to see all them clearly on a global map. The majority of the world's known threatened amphibians occur from Mexico south to northern Peru, and on the Caribbean islands. Most of the other important concentrations of globally threatened amphibians mirror the pattern for the other three groups, although eastern Australia and the southwestern Cape region of South Africa are also centres of amphibian threat. It should be emphasized that the paucity of data from certain parts of the world is probably severely underestimating concentrations of threatened amphibians, especially in the Albertine Rift, Eastern Himalayas, much of mainland Southeast Asia, Sumatra, Sulawesi, the Philippines, and Peru.
Photo 5.11
Gastrotheca ovifera (Endangered) is a species of marsupial frog. The eggs are carried on the female's back, which hatch into froglets. This species is restricted to the Venezuelan coastal range, where some populations appear to be in decline.
Photo: © Michael and Patricia Fogden.
An alternative perspective on the geography of threatened species is to measure numbers of threatened species relative to the overall numbers of a particular taxonomic group present. This is shown here for mammals, Western Hemisphere birds, freshwater turtles, and amphibians in Figure 5.10. The results are highly sensitive to change in areas with low overall species richness (e.g., the movement of a single species in the depauperate polar regions from one category of threat to another could make a large difference to the overall map), but nevertheless reveal some interesting additional patterns. For all four groups the proportion of fauna in danger of global extinction is high in island ecosystems such as the Caribbean, Madagascar, Sundaland, the Philippines, and New Zealand. For amphibians the map of relative threatened species richness largely parallels that for all species, whereas the relative distribution of threatened mammals and turtles is much more expansive. This covers threatened but species-poor areas of the temperate zone, such as California, the fringes of the Sahara, and central China; for amphibians the Argentinean Pampas and Asia Minor also stand out, as does northeast Canada for mammals.
Lack of comprehensive geographic and threat assessment for other species groups precludes the presentation of maps for these taxa. However, the availability of such data is likely to reveal many broad similarities with the patterns presented above for mammals, birds and amphibians, as well as some differences. For example, distribution patterns of threatened reptiles (in particular lizards) are likely to highlight the importance of some arid ecosystems. It is already known that some distribution patterns of threatened plants do not match those of most animal groups, some notable examples being the Cape Floral Region and Succulent Karoo of South Africa, and the deserts of the southwestern United States and northern Mexico. There are also very different patterns of threat among some freshwater groups. For example, the Mississippi drainage system is probably the global centre for threatened freshwater mussels. Patterns of threat in marine ecosystems will, of course, be completely different, and data on these patterns are still largely unavailable. However, the importance of the southern oceans in general, and the Tasman Sea and the southwestern Pacific around New Zealand for globally threatened seabirds can be seen in Box 5.1.
Figure 5.10 Threatened species richness relative to the overall number of species present in four taxonomic groups: clockwise from top left: mammals (marine species excluded); Western Hemisphere birds; amphibians; and turtles (at a half-degree resolution). Dark red colours correspond to higher richness, dark blue to lowest. Colour scale based on five equal-area classes.
Photo 5.12
Bastard Quiver Tree Aloe pillansii (Critically Endangered) is an example of one of the many threatened succulent plants found in the Succulent Karoo region of South Africa.
Photo: © Janice Golding.
Photo 5.13
The Mulanje Cedar Widdringtonia whytei (Endangered) is confined to the Mulanje massif in southern Malawi. This is an Alliance for Zero Extinction (AZE) locality.
Photo: © Craig Hilton-Taylor.
The ranges of globally threatened seabirds cover marine areas in the Economic Exclusion Zones of many countries, but also encompass large parts of the open oceans outside national sovereignty. For example, the highest densities of threatened birds at sea are found in international waters in southern oceans, with a particular concentration in the Tasman Sea and the southwestern Pacific around New Zealand. International cooperation is therefore required to conserve such species, many of which are threatened through incidental capture by commercial longline fisheries.
Figure Box 5.1 Density map of threatened seabirds across the southern oceans
Taken from BirdLife International (2004b)
As the spatial resolution of data on the geographic distributions of threatened species increases, so does the utility of these data for conservation (Collar 1993–4, 1996), but, unfortunately, the effort required to compile the data does as well. Nevertheless, the world's museums and herbaria represent a vast storehouse of such fine-scale geographic biodiversity data, and a number of initiatives are underway that suggest that these data will become increasingly available in the future (Bisby et al. 2000).
This said, synthesis of the numerous point data already available not only provides much finer resolution insight into the distribution of threatened species, but also provides a basis for establishing targets for site-scale conservation actions on the ground (see Section 8). The effort necessary to compile such data means that we are a long way from being able to show localities globally for all threatened species across multiple taxa. It is possible, however, to show important cross-sections of the data by trading these dimensions off against each other. Thus to allow mapping globally we have to compromise depth of coverage within taxa on the Red List. It is now possible to map localities (defined in Appendix 2e) for all threatened species within an individual taxon continentally (for birds, at least), and at a finer, regional scale, to map localities for all threatened species in the region of interest.
One important dataset concerns the distribution of Critically Endangered (CR) and Endangered (EN) species restricted to a single locality (www.zeroextinction.org). Figure 5.11 maps all sites known to hold the last remaining populations of a CR or EN mammal, bird, amphibian, or conifer species (as well as those reptiles assessed globally to date; cycads are pending). Broadly, most of these sites lie in the tropics, as one would expect, especially on islands. Interestingly, the map shows much stronger pattern in Latin America (Baja California, Caribbean, Tropical Andes, Atlantic Forest) and Africa (Cameroon Highlands, East African Highlands, Madagascar) than it does in Southeast Asia, where sites are scattered liberally across the continent.
A much broader approach is to compile data on all localities for all threatened species within given taxonomic groups. This clearly requires much greater investment in data collection and compilation, and so to date can only be presented as examples for specific geographic regions. One of the best examples comes from Africa, where the occurrence of all threatened bird species (Vulnerable (VU) species must exceed 10 pairs or 30 individuals at a single locality to be included in this dataset) has been mapped at the locality scale using the Important Bird Areas approach of BirdLife International (Collar and Stuart 1988; Fishpool and Evans 2001; Figure 5.12). Similar work has been completed in the Middle East, Europe, Asia, Canada, Mexico and the Andes, and is on-going in the Pacific, the rest of the Americas, Antarctica and marine areas. To date, some 4,032 sites holding threatened birds have been identified worldwide. Localities holding threatened bird species are highly clustered: in Africa, regions like the Mediterranean coast, Upper Guinea, the Cameroon highlands, the East and South African montane highlands, Madagascar, and the Indian Ocean Islands have particular concentrations of sites holding threatened birds. The Miombo-Mopane woodlands of Southcentral Africa hold a number of sites, albeit spread fairly far apart, and the Sahara-Sahel, Congo Forests, and Kalahari have very few localities hosting threatened species.
Figure 5.11
Map of localities (n=595) holding endemic, Critically Endangered (CR) or Endangered (EN) mammal, bird, turtle, crocodile, iguana, amphibian, and conifer species (source: the Alliance for Zero Extinction).
Figure 5.12
Map of localities (n=608) holding threatened birds in Africa (source: BirdLife International).
Expansion of the identification of individual localities holding threatened species beyond birds to cover all species on the Red List hits the most serious data limitations. However, new data from Madagascar, incorporating numbers of threatened mammals, birds, reptiles, amphibians, freshwater fish, and plants, demonstrates that this approach is possible for a wide range of taxa. Figure 5.13 represents a map of those 141 localities across Madagascar holding populations of these species. The map emphasizes the pattern of clustering seen for birds at the pan- African scale. Thus, a large number of localities holding threatened species are located in the country's eastern rain forest, western dry forest, and southwestern spiny forest, while the long-deforested central plateau holds very few (although the latter is also a region where many species extinctions have probably gone unrecorded).
Figure 5.13 Map of localities (n=141) holding the distributions of 754 globally threatened species in Madagascar, covering eight taxon groups: mammals, birds, amphibians, freshwater fish, reptiles, arthropods, gastropods and plants (preliminary unpublished data provided by Zo Lalaina Rakotobe, Luciano Andriamaro, Harison Rabarison, and Harison Randrianasolo).
Photo 5.14
The Aquatic Tenrec Limnogale mergulus (Endangered) requires clean and fast flowing water and is therefore threatened by the increasing siltation of the rivers, due to soil erosion following deforestation in Madagascar.
Photo: © P.J. Stephenson.
Photo 5.15
Madagascar Teal Anas bernieri (Endangered) has a very small population that is declining rapidly due to habitat loss and hunting.
Photo: © Frank Todd.
Photo 5.16
Pachypanchax sakaramyi (Critically Endangered) is an endemic Madagascan killifish known only for certain from a few small puddles fed by a leaking water tap, following the diversion of the headwaters of the Sakaramy River for domestic use by local people.
Photo: © Paul Loiselle.
The overall message from this section is that threatened species are distributed highly unevenly around the surface of the planet. Within this, five other key conclusions emerge:
Patterns of threatened species distributions are remarkably congruent between taxa analysed. Differences are driven by underlying rangesize distributions among taxonomic groups (e.g., birds tend to have larger range sizes than amphibians), and by ecological limitations of specific taxa (e.g., birds are better able to disperse over saltwater than amphibians).
Most threatened species occur in the tropics, especially on mountains and islands. In the marine realm, the ‘coral triangle’ of the western Pacific and eastern Indian Ocean holds the most threatened species in most taxa.
The uneven distribution of threatened species means that most known threatened species occur in a few countries.Australia, Brazil, China, Indonesia, and Mexico hold particularly large numbers of threatened species and endemics.
Most threatened species assessed to date are terrestrial. The smaller numbers of threatened species in marine and freshwater systems are artefacts of lack of assessment, and preliminary indications suggest that freshwater species, in particular, may actually be proportionately more seriously threatened than terrestrial species.
Decisions regarding resolution of mapping involve trade-offs regarding the geographic extent of mapping, the breadth of taxonomic coverage, and the depth of coverage through the Red List Categories.
Photo 5.17
The Seychelles Scops-owl Otus insularis (Endangered) is endemic to Mahé in the Seychelles. Previously Critically Endangered, this owl's status has recently improved as its favoured upland forest habitat has increased in extent over the last 40 years. However, the population is still extremely small and the species is susceptible to introduced predators such as Black Rat Rattus rattus and Barn Owl Tyto alba.
Photo: © Dave Currie.