Africa occupies about one-fifth of the global land surface and contains about one-fifth of all known species of plants, mammals, and birds in the world, as well as one-sixth of amphibians and reptiles (Siegfried, 1989). This biodiversity is concentrated in several centers of endemism. The Cape Floral Kingdom (fynbos), which occupies only 37,000 km2 at the southern tip of Africa, has 7,300 plant speciesof which 68% occur nowhere else in the world (Gibbs, 1987). The adjacent Succulent Karoo biome contains an additional 4,000 species, of which 2,500 are endemic (Cowling et al., 1998). These floristic biodiversity hotspots both occur in winter rainfall regions at the southern tip of the continent and are threatened particularly by a shift in rainfall seasonality (for instance, a reduction in winter rainfall amounts or an increase in summer rainfall, which would alter the fire regime that is critical to regeneration in the fynbos). Other major centers of plant endemism are Madagascar, the mountains of Cameroon, and the island-like Afromontane habitats that stretch from Ethiopia to South Africa at altitudes above about 2,000 m (Mace et al., 1998). Montane centers of biodiversity are particularly threatened by increases in temperature because many represent isolated populations with no possibility of vertical or horizontal migration. Several thousand species of plants alone are potentially affected.
The broad patterns of African zoogeography also are climatically linked, but the location of concentrations of biodiversity and endemism, at least in the higher animals, is located in the savannas and tropical forests. World antelope and gazelle biodiversity (more than 90% of the global total of 80 species) is concentrated in Africa (Macdonald, 1987). Changes in climate of the magnitude predicted for the 21st century could alter the distribution range of antelope species (Hulme, 1996).
This biodiversity forms an important resource for African people. Uses are consumptive (food, fiber, fuel, shelter, medicinal, wildlife trade) and nonconsumptive (ecosystem services and the economically important tourism industry).
For a sample of 39 African countries, a median 4% of the continental land surface is in formally declared conservation areas in southern Africa (MacKinnon and MacKinnon, 1986). The fraction of landscape that is conserved varies greatly between countries (from 17% in Botswana to 0% in four countries), as does the degree of actual protection offered within nominally conserved areas (MacKinnon and MacKinnon, 1986). A very large fraction of African biodiversity occurs principally outside of formally conserved areas (especially in central and northern Africa), as a result of a relatively low rate of intensive agricultural transformation on the continent. This will no longer be true if massive extensification of agriculture and clearing of tropical forests occurs in the humid and subhumid zones, as is predicted to occur in the next century by some land-cover change models (Alcamo, 1994). Patterns of human pressure, including grazing by domestic stock, also will be altered and intensified by climate change. Land-use conversion effects on biodiversity in affected areas will overshadow climate change effects for some time to come.
In the medium term (~10-20 years), biodiversity of indigenous plants and
animals in Africa is likely to be affected by all of the major environmental
changes that constitute climate change. These include changes in ambient air
temperature, rainfall and air vapor pressure deficit (which combine to cause
altered water balance), rainfall variability, and atmospheric CO2.
Africalike the other continents, though perhaps to a greater degreeis characterized by ecosystem control through disturbance, such as fire (Bond and van Wilgen, 1996) and grazing regimes. Changing disturbance regimes will interact with climate change in important ways to control biodiversityfor instance through rapid, discontinous ecosystem "switches." For example, changes in the grazing and fire regime during the past century are thought to have increased woody-plant density over large parts of southern Africa. Ecosystem switches are accompanied by drastic species shifts and even species extinction. Subtle changes in species composition of rich ecosystems such as forests will impact biodiversity resources. A significant reduction in rainfall or increase in evapotranspiration in Angola would threaten the Okavango delta wetland in Botswana. Much larger scale ecosystem switches (e.g., savanna to grassland, forest to savanna, shrubland to grassland) clearly occurred in the past (e.g., during the climatic amelioration dating from the last glacial maximum), but diversity losses were ameliorated by species and ecosystem geographical shifts. The geographical range shifts recquired to preserve biodiversity into the future will be strongly constrained by habitat fragmentation and cannot realistically be accommodated by a static nature reserve network with the low areal coverage evident in Africa.
Theory required to predict the extent and nature of future ecosystem switches and species geographical shifts in Africa is lacking, and case studies are few. The response of major vegetation types to changes such as rising atmospheric CO2 are almost unstudied, although early evidence (Midgley et al., 1999; Wand et al., 1999) suggests, for example, that these responses may increase WUE in grass species significantly, which may increase grass fuel load or even increase water supply to deeper rooted trees. Recent analysis of tree/grass interactions in savannas suggests that rising atmospheric CO2 may increase tree densities (Midgley et al., 1999); this kind of ecosystem switch would have major implications for grazing and browsing animal guilds and their predators. For southern Africa, between one-quarter and one-third of current reserves were predicted to experience a biome shift (a major change in the dominant plant functional types) under the equilibrium climate resulting from a twice-preindustrial CO2 concentration (Hulme, 1996). In South Africa, increased aridity in the interior Bushmanland plateau will introduce a desert-like environment to the country (Rutherford et al., 1999). Analysis for South African conservation areas (Rutherford et al., 1999) shows potentially large losses of plant species diversity in this semi-arid region with low landscape heterogeneity.
Thus, the vegetation and animal communities that many reserves aim to conserve will no longer be within their preferred bioclimatic region. Migration of animals to conserved areas with more suitable climate (if these exist) will be constrained by fragmentation of intervening ecosystems and potentially hostile landscapes. The required rate of migration may be too rapid for unassisted movement of most plant species, especially over relatively flat landscapes (Rutherford et al., 1995). Without adaptive and mitigating strategies, the impact of climate change will be to reduce the effectiveness of the reserve network significantly, by altering ecosystem characteristics within it and causing species emigrations or extinctions.
At particular risk of major biodiversity loss are reserves on flat and extensive landscapes, those in areas where rainfall regime may change seasonality (e.g., the southern Cape), those where the tree/grass balance is sensitive to CO2 conditions, and those where the fire regime may be altered. Species most at risk are those with limited distribution ranges and/or poor dispersal abilities, habitat specialists (soil specialists in the case of plants), and those that are responsive to specific disturbance regimes.
Mitigation and adaptation strategies will be greatly strengthened by a risk-sharing approach between countries, which could attempt to share the burden of conserving critical populations in a collaborative way. Part of this risk-sharing approach could include transboundary nature reserves, where this is appropriate for increasing connectivity in areas projected to change significantly. The corridor approach within and between countries would have the added benefit of increasing reserve resilience to current climate variability and would increase attractiveness to the tourism industry. Economic incentives, however, may differ across geographic scales. For the moist tropical forests of Masoala National Park in northeast Madagascar, economic incentives favor conservation at local and global scales, although logging provides more profit at a national scale (Kremen et al., 2000).
A high degree of uncertainty is associated with predictions of the biodiversity effects of climate change. No systematic analysis of mechanisms of ecosystem switches, or areas exposed to them, has been carried out. Although fire and atmospheric CO2 seem to be important determinants of ecosystem structure and function, little research is available to predict how these factors will interact with other environmental changes. The effects of CO2 on grass water use may be an important mitigator of negative effects on productivity for grazer guilds in much of subtropical Africa. Effects of shifts of disease-prone areas on animal populations are unstudied.
Other reports in this collection