The Regional Impacts of Climate Change

	Other reports in this collection

2.3.1.4. Forest and Woodland Ecosystems

Forests in Africa are of great socioeconomic importance as sources of timber, fuel, and many nonwood products, as well as for the protection of water resources. Ecologically, they serve critical roles in water, carbon, and nutrient cycling. The impacts of climate change on forests at the continental scale will be assessed in very broad terms using biome distribution models.

Figure 2-6: Distribution of forest, core areas, and gradients of decreasing biotic diversity in tropical Africa. Core areas are believed to approximate sites of forest refugia at the time of the last world glacial maximum (18,000 BP) (Hamilton, 1988).

In geological time scales, forest boundaries fluctuated a great deal during the Pleistocene epoch (Sayer et al., 1992); the forests of Africa even now are considerably more extensive than they were during the most recent high-latitude glacial advance about 18,000 years ago. Environmental conditions in tropical Africa at about 18,000 before present (BP) are quite well known, thanks to a large number of pollen and plant microfossil studies (see Hamilton, 1988, for more details). During the severe arid period around 18,000 BP, core areas (centers of biotic diversity) were the main centers of forest survival. Two such principal core areas are located in Cameroon/Gabon and eastern Democratic Republic of Congo (DRC) (formerly Zaire); other, less-diverse core areas are in west Africa and near the east African coast (Sayer et al., 1992). The core areas are not only rich in numbers of species and endemics but also are centers of distribution of disjunct species. Some of these species are unlikely to be able to disperse from core area to core area without continuous forest cover. For example, gorillas are disjunctly distributed across the Zaire basin (Figures 2-6 , 2-7 and 2-8). Although the forests between the two populations seem suitable for the species, some explanation is needed regarding how their ranges became fragmented. A likely explanation for the gorilla and other obligate forest species is that their ranges became fragmented as a result of forest retraction at times of aridity and that these species subsequently have been slow to expand their range to include all potentially suitable habitat. Thus, animal species may not adapt fast enough to rapid changes in climate.

Figure 2-7: Forest refugia during arid periods in central Africa: A) Central refuge; D,C,G) Cameroon/Gabon refuge (D, Niger section; C, Cameroon section; G, Gabon or Ogoowe basin section); B) southern Zaire basin refuge; E) north Angola refuge; F) southern scarps of Zaire basin; and H) Lunda Plateau (Kingdon, 1980).

Figure 2-8: The distribution of gorillas in Africa (Harcourt et al., 1989).

Current vegetation distribution can be studied using biome distribution models. Biome distribution models such as MAPSS (Neilson, 1995) and BIOME3 (Haxeltine and Prentice, 1996) simulate the distribution of potential global vegetation based on local vegetation and hydrological processes and the properties of plants. They simulate the mixture of life forms (such as trees, shrubs, and grasses) that can coexist at a site while in competition with each other for light and water. These models simulate the maximum carrying capacity, or vegetation density (in the form of leaf area), that can be supported at the site under the constraints of energy and water. A change in leaf area can be interpreted as a change in overall carrying capacity or standing crop of the site, regardless of whether it is potential natural vegetation or under cultivation. This carrying capacity potential is the basis for application of these models to all of Africa to indicate possible shifts in agricultural or vegetation potential.

MAPSS and BIOME3 outputs were generated (see Annex C) using several GCM scenarios, with and without CO₂ effects and aerosol emissions. Table 2-1 summarizes changes in areal coverage for four main biome types in Africa (the shrub/woodlands biome was combined with grasslands). The models indicate a net shift from more arid biomes (low leaf area index-LAI) to more mesic biomes (higher LAI) for the OSU, GFDL, and UKMO scenarios. Exact percentages of change and where this change would occur are highly uncertain because the models indicate potential, not actual, vegetation.

The biome models are able to capture some overall divisions (e.g., rainforest versus woodland versus arid), but they do not yet capture fine-scale detail for Africa. This limitation is a result of the quality of the climate data used as inputs, coarse soils information, and the nature of the models-which were primarily designed to describe vegetation in temperate climates.

Biome distribution models simulate only static vegetation (i.e., equilibrium vegetation at a future date). Factors that affect vegetation dynamics, such as competition and disturbance (fire, herbivory), are not considered. These factors have not yet been incorporated in global vegetation models (but see Woodward and Steffen, 1996, for plans and developments in this area). An important factor to consider with regard to whether a vegetation type can respond as simulated is migration. Migrating species would require corridors of unmanaged land. The fragmented nature of remaining African vegetation (outside the rainforests) would make vegetation responses to climate change difficult. The destructive aspects of fires also would reduce migration. The dynamics of savannas and woodlands (such as miombo) are strongly linked to fires, so likely changes in fire intensity and frequency will have unknown consequences on vegetation.

Some progress is being made in the development of models of vegetation dynamics that include the effects of fire and herbivory for African vegetation (e.g., Menaut et al., 1991; Van Dalaan and Shugart, 1989; Desanker, 1996). However, these models have to be widely tested and validated before they can be used to evaluate impacts of climate change at the broad scale. Results of biogeochemical modeling in savannas using the CENTURY model of Parton et al. (1992) are discussed in Section 2.3.1.5.