Global Mammal Conservation: What Must We Manage?

See allHide authors and affiliations

Science  22 Jul 2005:
Vol. 309, Issue 5734, pp. 603-607
DOI: 10.1126/science.1114015


We present a global conservation analysis for an entire “flagship” taxon, land mammals. A combination of rarity, anthropogenic impacts, and political endemism has put about a quarter of terrestrial mammal species, and a larger fraction of their populations, at risk of extinction. A new global database and complementarity analysis for selecting priority areas for conservation shows that ∼11% of Earth's land surface should be managed for conservation to preserve at least 10% of terrestrial mammal geographic ranges. Different approaches, from protection (or establishment) of reserves to countryside biogeographic enhancement of human-dominated landscapes, will be required to approach this minimal goal.

Research on population and species extinctions shows an accelerating decay of contemporary biodiversity. This pressing environmental problem is likely to become even worse in coming decades (13). Although impacts of human activities are global in scope, they are not uniformly distributed. The biota of certain countries and regions can be identified as being most at risk, having both exceptionally high richness and endemism and exceptionally rapid rates of anthropogenic change. Because resources for conservation are limited, ecologists must provide managers and politicians with solid bases for establishing conservation priorities (4) to minimize population and species extinctions (5), to reduce conservation conflicts (6, 7), and to preserve ecosystem services (8).

Even for charismatic taxa, we lack a global view of patterns of species distributions useful for establishing conservation priorities. Such a view would allow evaluation of the effort required, for example, to preserve all species in a given taxon. It would also be relevant to setting global conservation goals such as protecting a certain percentage of Earth's land surface (9). More restricted approaches such as identifying hot spots and endemic bird areas have called attention to relatively small areas where large numbers of species might be protected (1013). For instance, recently the number of vertebrate species that lack populations within major protected areas was estimated (12). But now more comprehensive analyses are possible.

Here we conduct a global examination of mammal distributions to evaluate conservation priorities based on (i) range size distribution, (ii) global patterns of species richness, (ii) political endemism (i.e., the proportion of species restricted to one country), (iv) the minimum area required to preserve one population or 10% of the range of each species, and (v) conservation conflicts in priority areas.

We created maps for 4795 mammal species, excluding only marine species, from the literature (14, 15). To evaluate the minimum area required for preserving these mammal species, we compared “minimal” and ”conservative” preservation criteria. Under the minimal criterion, cells required to have appropriate management to preserve all mammal species in at least one 10,000-km2 cell were selected. Under the conservative criterion, enough cells were selected to preserve a minimum of 10% of the range of each species. Using a percentage criterion was judged better than selecting a number of cells, because we are only dealing with conservation of species here. The much more difficult and possibly more important issue of population conservation to maintain ecosystem services (2) is only partially considered and obviously would require even more extensive management. Our selection of cells includes the complete distribution of many species with a species range (SR) equal to or smaller than 10,000 km2, and a large percentage of the SR of species with a range smaller than 50,000 km2. A database of cells and the species found in each was entered into the MARXAN Reserve Design program (version 1.8.2) (16) to produce 250 scenarios for both minimal and conservative preservation criteria (15). Each scenario was a result of MARXAN's simulated annealing algorithm set to produce an optimal solution based on 10,000 iterations. We used the best solution (the one needing the minimum number of cells) to represent our global conservation management network (16). We used a fractional crop cover data set to estimate the proportion of each cell that is occupied by cropland (17). Spatially referenced human population data from the Center for International Earth Science Information Network [CIESIN (18)] were used to determine the population density for each cell in the global reserve network.

Species with small geographic ranges are more vulnerable to human impacts (and thus to extinction) than are widespread ones, and the number of those restricted-range species is positively related to the number of sites required to preserve global mammalian diversity. Although the geographic distributions of land mammals vary from very small (<10 km2; one cell) for some island species to very large for species such as the wolf (Canis lupus, >49 million km2; 4900 cells), most (76%) species have a SR smaller than 1 million km2 (100 cells) (Fig. 1). Rare species (n = 1198), defined here as those comprising the first quartile of the frequency distribution of geographic ranges, have a SR smaller than 24,000 km2 (less than three cells, roughly half the size of Costa Rica). There are more rare species in Asia, followed by Africa, South America, North America, and Australia. Similarly, 32% (1531) of the species have a SR smaller than 50,000 km2 (five cells), which is a threshold used with other factors to determine species endangerment by international conservation agencies such as the World Conservation Union (19). An unexpected result is that only 8% of mammal species are exclusively found in hot spots [sensu (11)], 62% are shared between hot spots and other regions [i.e., “cold spots” sensu (13)], and 30% are restricted to cold spots. Similarly, the ranges of 95% of all species intersect with at least one reserve in the 2004 IUCN/UNEP (World Conservation Union/United Nations Environment Programme) World database of protected areas.

Fig. 1.

Most mammal species have relatively small geographic ranges (<400,000 km2), encompassing 20% or less of the continent where they occur. Such limited geographic ranges tend to make those species relatively prone to extinction.

As expected on the basis of biogeographic theory and patterns in other taxa, distribution of mammal species richness is very heterogeneous, with regions of low and high diversity on each continent and higher richness at lower latitudes (Fig. 2). As a result, the number of species in a single cell varies from 10 to 257. Most cells throughout the world have relatively few species (<100), especially in large regions in northern Africa and Asia, and nearly all of Europe and Australia. Unexpectedly, only four regions—Central America, the Andes-Amazonia in South America, east-central Africa, and Southeast Asia—have very rich cells, containing 200 or more species. South America has by far more of these cells.

Fig. 2.

Patterns of mammal species richness in six major regions of the world. The abscissa shows number of cells, and the ordinate shows species richness. Most diverse regions are found in South America and Africa. Marked cells indicate priority areas for the maintenance of 10% of the geographic ranges of all mammal species.

The threat to the almost 40% (1900) of politically endemic (5) mammals is at least partially negatively correlated with economic development. Developing nations often lack resources for conservation. Centers of political endemism, with 5% or more of the world's endemic mammals each, are Australia, Indonesia, Mexico, Brazil, the United States, Philippines, Madagascar, China, and Papua New Guinea. Most countries, and most megadiverse countries, are underdeveloped, and endemic species are concentrated in those countries. Some 47% (906) are found in countries like Iran, ranking below the top 100 countries in PPP (Purchase Parity Power; <U.S. $5900); only 18% of mammal species are politically endemic to industrialized countries (PPP > U.S. $11,000) (20).

The minimal preservation criterion (all species in at least one 10,000-km2 cell) requires the management of 668 cells (6,680,000 km2), ∼4.2% of Earth's ice-free land surface. In contrast, the more conservative criterion (cells to represent 10% of the geographic range of all species) requires managing at least 1702 cells (17,020,000 km2), accounting for 11% of the ice-free land surface. Cells in the conservative scenario were located in Asia (589), Africa (349), North America (299), South America (220), Europe (126), and Australia (119). Many mammal species (3293, 68%) were represented in very few priority cells (<10); 6% (290 species) were found in more than 100 cells, and the range was from 1 to 404 cells (Fig. 3). The total number of species occurrences in these cells, which could be used as a very conservative estimate for the number of populations [Mendelian populations or demographic units (2)] of mammals, is 116,103.

Fig. 3.

Frequency distribution of the number of cells occupied by mammal species in the conservative criterion, in which at least 10% of the range area of each species is protected. Most of these species occur in fewer than 40 100-km by 100-km cells, equivalent to the area of California (USA) or three-quarters the area of France.

Practically, it is important to carry out a sensitivity analysis to determine the degree of substitutability of cells; that is, if one becomes uninhabitable, how readily can a substitute be found. It is highly unlikely that all selected cells will retain high conservation value because human impacts and removing a degraded cell can affect the value of others in MARXAN scenarios. Hence, strategic conservation investment would involve back-up or next-best site sets as well as the priority set. More than 93% (15,057) of cells were selected in at least 1 of 250 scenarios for the 10% preservation criterion, and 135 cells (<1%) were irreplaceable, that is, selected in all scenarios (Fig. 2). Irreplaceable cells were located in all continents; most were in Asia (51, 37%), followed by North America (31, 23%), South America (20, 15%), Africa (19, 14%), Australia (10, 7%), and Europe (4, 4%). This is a favorable outcome, because 95% of the cells can be replaced by other cells without loss of conservation value, which gives the opportunity for strategic conservation planning. On the other hand, 1225 (72%) and 120 (89%) of the 1702 and 135 priority or irreplaceable cells in the best solution, respectively, intersect any one of the IUCN protected area polygons, indicating a relatively good correspondence between priority cells and protected areas.

To evaluate the extent to which a few species are driving the selection of priority cells, we quantified the difference in number of grid cells needed to achieve the conservative goal as compared to the number needed to protect 10% of the range of species endangered according to the IUCN (21). Only 48% (814) of cells would be required to represent those species, and those cells are found in North America (32%), Africa (27%), Asia (18%), South America (13%), Australia (8%), and Europe (2%). Similarly, we determined that to represent all mammals except rodents, the most diverse order, required 5% fewer cells. Randomly reducing the number of species to be protected at 10% intervals showed very slow reductions in the number of cells. Dropping 10, 20, 50, and 90% of the species represented a reduction of >1, 2, 7, and 21% of cells in the conservative (10%) representation goal.

Although land area under agriculture and human population density by themselves might seem to be crude estimators of anthropogenic impact, they actually are good indicators of overall biodiversity loss and conservation conflict, defined as the overlap of human activities and priority areas for conservation (2226). We found a positive relation between species richness and human population density within the conservative, 10% criterion cells (Fig. 4A). One alarming result that requires immediate attention is that ∼80% of the land area that must be managed under this conservative criterion has been affected to some degree by agriculture (Fig. 4B). Indeed, 20% of such cells have lost from 26 to 100% of their natural vegetation to agriculture, reducing the value of each for conservation (Table 1).

Fig. 4.

Conservation conflict as represented by the relation between mammal species richness and human population density (A), and agriculture (B), in the priority cells. (A) A Michaelis-Menten type of curve was adjusted by a nonlinear method, and the r2 (proportion of variance accounted for by the model) was calculated for each continent. Human population and number of mammal species are more strongly associated in Europe (r2 = 0.77), Africa (r2 = 0.59), and Asia (r2 = 0.53) than in North America (r2 = 0.29), South America (r2 = 0.19), and Australia (r2 = 0.09). (B) A simple linear fit yields a low value of r2 (0.09), indicating a lack of relation between agriculture and number of mammal species.

Table 1.

Percentage of each priority cell for conservation covered by cropland (17) in six regions of the world. This is a measure of conservation conflict. Many cells have less than 25% conversion to cropland, suggesting opportunities for ameliorating conflict through countryside biogeography.

Cells Africa Asia Australia Europe North America South America Total
None 20 14 55 6 36 1 20
1–25 70 58 39 35 41 95 59
26–50 6 13 5 29 14 4 11
51–75 4 11 1 22 6 7
76–100 4 8 3 3

Our results have clear implications for conservation. First, the large number of mammal species with restricted geographic ranges calls forth complex conservation scenarios because those species are intrinsically vulnerable to human impacts and stochastic extinction, and many are politically endemic. The high frequency of restricted ranges reinforces the obvious point that a larger and more dispersed selection of sites is required to preserve global mammalian diversity than would be required if range size distribution were skewed in the other direction. As the human population and its impacts have escalated, being politically endemic with a restricted geographic range in developing nations (e.g., orang-utan, Pongo pygmeus, in Indonesia) could, in most cases, considerably increase extinction risk because conservation legislation and resources are lacking. In contrast, the many endemic species in Australia are now subject to the kinds of intense management efforts that only rich nations are able to afford. Further complicating the conservation task is that even a well-known taxon such as mammals is more speciose than previously thought. There are now 10% more mammal species than were estimated a decade ago (27), and the total number is certain to be higher.

Second, our study supports previous ones showing that a large fraction of Earth's surface is important for the conservation of species diversity (9, 11, 12, 25). This is especially true given that we focus only on mammals—the cells that would be selected for, say, butterflies or freshwater fishes would likely be quite different (9). Obviously, it is very important to designate and manage reserves from which human activity is excluded or strongly restricted. Many species and populations such as the mountain gorilla (Gorilla gorilla) in Africa could not now survive without effective reserves. Conservation by protected areas, however, although effective and necessary, cannot be the only strategy. Even under optimistic assumptions, managing just 4% of Earth's surface, as in our minimum protection scenario, would be a gigantic task, especially when considering both high levels of conservation conflict throughout the world and gaps in the representation of species in protected areas.

In some places, the required habitat types or resources could be protected without formal reserves, through conservation finance approaches or through other cultural and political mechanisms. This work and related studies throw into sharp relief the importance of areas outside of protected parks and reserves for the maintenance of mammalian diversity. This is the domain of countryside biogeography (8). Many species can and do survive in regions with different degrees of human impact. For example, research in the Las Cruces region in southern Costa Rica indicates that a large percentage of the mammals (and other taxa) can survive in a region with ∼9% of its original forest cover remaining (28). Similar patterns have been found throughout the world, even for populations of very large species such as tigers in India and Nepal (29). From the viewpoint of both ecosystem services and biodiversity preservation, we now need to start managing the whole planet better, as Vitousek et al. (30) point out. In some sense, “we” already manage the whole planet. However, the problem of planetary management, paying careful attention to “priority” grid cells and political endemism for a diverse selection of taxa, is especially daunting because of the great differences in the ability of nations to protect organisms within their borders. That, and the scale of the problem, mean that an unprecedented international effort will be needed—one requiring the development of both new attitudes and institutions (31). Developing those in a politically sensitive and cost-effective way is perhaps the major challenge for conservation biology.

Supporting Online Material

Materials and Methods


Species list

References and Notes

View Abstract

Stay Connected to Science

Navigate This Article