WHITE'S 1983 VEGETATION MAP OF AFRICA: A CONTRASTING APPROACH

A useful example of a more structured and objective approach is that of Frank White and his collaborators. White (1983, 1993) published a vegetation classification and map for continental Africa, based on the concept that physiognomic data, and in particular the cover and height of the vegetation, are fundamentally different from chorological data on patterns of species distribution. His approach was to study these two aspects independently, starting with physiognomy. White developed an intentionally simple system in which 17 major vegetation types are recognized in 5 categories. For example, forests are defined as all vegetation comprising a continuous stand of trees at least 10m tall with interlocking crowns.

Bushland (formation buissonnante) includes open stands of bushes 3-7 m tall, with a canopy cover of 40% or more, and woodland (forÊt claire) is defined as open stands of trees at least 8 m tall, with a canopy cover of 40% or more, and with a ground layer usually dominated by grasses and other herbs, as can be seen in illustrative forest profiles prepared by Pratt et al. (1966).

Didiereaceae thickets (fourrés à Didiereaceae), and indeed all other vegetation types, are an ecological phenomenon that reflects the environmental milieu at a given place at a particular point in time (see Emberger, 1942, 1955, 1971; Cornet and Guillaumet, 1976). The most important overall factor in determining physiognomy is climate, although local conditions such as soils or water availability can also be important (Basse, 1934). It is important to note that grasslands (formations herbeuses) were defined by White (1983) specifically as land covered with grasses and other herbs, with woody plants covering no more than 10% of the ground. Several authors have argued strongly against the use of imported terms such as prairie, savanna, steppe and pseudosteppe in Africa (see Chapman and White, 1970; Descoings, 1973, 1978; White, 1983). For example, savanna has been applied in so many ways that it is highly ambiguous (see also Eiten, 1986, 1992); steppe should be restricted to the characteristic grasslands of eastern Europe (cf. Eiten, 1992); and prairie refers to a type of vegetation that was once widespread in central North America. White's (1983) classification accommodates these and other forms of vegetation without having to use such problematical terms (but see Lawesson, 1994, for some critical comments), and we feel that it would be appropriate to abandon them definitively when discussing the vegetation of Madagascar.

12 species of Diospyros (Ebenaceae) fruiting simultaneously at Nosy Mangabe Special Reserve

In contrast to vegetation structure (physiognomy), the floristic composition of vegetation at a given place -- i.e., the assemblage of individual species that are present -- is a consequence of evolutionary factors, as well as the dispersal and migration of species through time. In other words, the plants that one sees at a particular site reflect historical events over evolutionary time (Emberger, 1942, 1955, 1971; Cornet and Guillaumet, 1976).

The geographic distribution of plant species is not entirely random. Clear patterns exist, and they can be detected by careful analysis of the distribution patterns of a large numbers of species. White (1976, 1979, 1983, 1993) used this kind of chorological approach in Africa. For example, the Guineo-Congolean regional center of endemism (Zone I on White's map) comprises about 8,000 spp, of which about 80% are endemic to the region; there are also 9 endemic plant families and 1/4 of the genera are endemic (White, 1983, 1993). It is important to note that all of these regions, or phytochoria, are based exclusively on chorological data.

Once defined, however, these regions almost always comprise areas with very similar bioclimatological characteristics, as illustrated by clima-diagrams from various sites within the Sudanian regional center of endemism, which spans from Senegal to Ethiopia (White, 1983). White mapped his basic physiognomic categories (forest, bushland, thicket, etc.) onto the chorological zones, which serve to subdivide the vegetation types into cohesive regions on the basis of chorology.


Previous Page--Contents--Next Page