In a paper entitled ‘Africa – the odd man out’ Richards (1973) pointed out that of the three main regions of tropical rain forest he recognised, namely American, African and Indo–Malaysian, the African rain forest stands out as distinct from the others. He took this view largely on the comparative poverty of its flora, the wider distribution of African plants, and the poor representation of certain plant groups such as palms, orchids and trees of the family Lauraceae. In addition, Africa is relatively poorly endowed with some life forms such as lianes and epiphytes. He attributed the uniqueness of tropical Africa to a number of possible factors. In the first place, differences may be due to climate since the uniformly high rainfall, humidity, temperatures, and absence of a regular dry season that occur in parts of the other two tropical regions are evidently rare or lacking in Africa. A distinct dry season, often of over two months, is nearly always present, and even in the central Congo basin there is usually at least one month when rainfall is less than 100 mm. Secondly, the impact of human beings on vegetation has been longer and more sustained in Africa than elsewhere, so much so that the presence of any truly primeval or so-called virgin forest is very questionable. However, it is unlikely that present day climatic factors and human interference can wholly account for the unique position of African rain forest, and historical reasons must also be sought. Thus it seems probable that the vicissitudes of climate, recorded as arid and pluvial periods in African lake sediments, have been much greater than those in the other two regions in question.

The forest zone of West and Central Africa comprises the coastal and adjacent inland regions bounded the semi-deciduous forests in the west and the equatorial forests in central Africa and the Congo basin. Sedimentary plains, developed mostly on weathered sandy materials, lie along the coastal stretches and cover vast areas of the Congo basin. Behind the coast the plain rises gradually to hills and plateaus of much lower elevation than those of the highlands of East Africa. Two great rivers, the Niger and the Congo, which discharge huge volumes of fresh water to the Atlantic Ocean, are major contributors to the hydrological cycles of the rain forests ecosystems of the Guinea–Congo Region. The Niger originates from the forested highlands of Guinea and discharges its waters into the Bight of Benin by way of a large delta in southern Nigeria. The Congo basin occupies an immense area of 750 000 km2, surrounded by Pre-Cambrian uplands. The alluvial floor of the saucer-shaped basin is flat, and marshes and swamps comprise a large proportion of the total area. The highlands and plateaus along the rim are low to the west and north and higher to the south. To the east, they merge with the mountains of the Great Rift Valley of East Africa (Gann & Duignan 1972; Hance 1975; Grove 1978; Hamilton 1989).

This chapter sets out to give a historical overview of the African rain forest from its origins, towards the end of the Cretaceous period. The areas around the Gulf of Guinea, in particular from Ivory Coast to Nigeria and especially Cameroon, Gabon and Congo, appear to have been already occupied at this time by wet tropical forest formations mainly composed of Angiosperms which were then becoming established. In the course of the Tertiary period the combined effect of the equator being situated further north than now and the development of the Antarctic ice cap favoured the development of wet tropical conditions over a large part of North Africa which in turn led to the extension of tropical forest to various sites on the shores of the Tethys Sea. There were probably at this time common taxa and similar vegetation patterns stretching from the Gulf of Guinea to the Tethys Sea.

Towards the end of the Tertiary, the equator reached its present position and the northern hemisphere ice caps appeared, and these phenomena resulted in the disappearance of the forest formations spread across the north of Africa, and the concentration of these formations near the equatorial zone around the Gulf of Guinea and in the Congo–Zaïre basin. From 800 000 years ago onwards the marked glacial variations at middle and high latitudes in both hemispheres, with a periodicity of about 100 000 years determined by the orbital variations of the earth around the sun, lowered temperatures in equatorial areas and brought arid climates at times of maximum glacial extension. The most arid periods resulted in the fragmentation of the forest cover, and the forest biotopes and their biodiversity were preserved in a series of refugia. The lowering of temperatures also resulted in the extension of montane flora to low altitudes, with migration of montane flora and fauna between main mountain ranges. These compounded phenomena of isolation and migration, probably involving genie exchange, must have resulted in numerous speciation phenomena. Subsequently, such montane flora or fauna became isolated on mountain areas during periods of maximum warming, in the last instance in the course of the Holocene, when a vast forest cover became re-established from Guinea westwards, and to the East as far as the Lake Victoria area. The phases of maximum fragmentation, which appear to have been connected with only the coldest periods – in the last instance during the second part of isotopic stages 6 (from c. 160 to 130 000 years) and 2 (from c. 24 to 12000 years BP) – relate to less than 10% of the last 800 000 years, and the phases of maximum forest extension would likewise appear to be less than 10% of the period. The remaining 80–90% of the time relates to ‘intermediate situations’ which varied from period to period, and these intermediate extension situations seem to have been the norm over the larger part of the Quaternary, rather than the present situation which is closer to a situation of maximum extension.

The current state of Ghana's forest is summarised. Considerable changes have occurred in the last decade, since Hall & Swaine's account and classification, due mainly to fire and logging. The requirements and potential for sustainable forest use are explored through a summary of patterns of regeneration, and of local and national distribution of individual species.

Incisive indices of forest quality and condition are vital to good forest management. Various forest quality indices, summarising different properties of the plant community, are examined. These indices gloss over the statistically noisy behaviour of single species in small forest areas. The indices are: Forest Type – Hall & Swaine's forest ordination and classification; a Pioneer Index (PI) revealing the balance of ‘regeneration guilds’; a Genetic Heat Index (GHI), based mainly on the rarity value (Star rating) of all forest species, highlighting ‘hotspots’; and an Economic Index (EI) based on the concentration of common species (‘reddish Stars’) threatened by exploitation. Guild and Star are defined for all species and encapsulate trends of local and of global distribution and ecology. The national and local patterns and response to disturbance of the indices derived from the representation of these various guilds and stars are discussed.

Scale is crucial to all discussions. A strictly hierarchical model of forest ecology/biogeography is less suitable than a continuum-of-significant-scale, and non-hierarchical model. For instance, refugia are usually perceived as discrete biogeographical units. However, major biological ‘hotspots’, which are often described as refugia and attributed to Pleistocene climatic variation, differ only in position along a continuum of scale from mini-refugia as small as individual plants. The biogeographic Dahomey gap has much in common with a canopy gap, with scale as the main distinction.

There are conspicuous trends across Ghana's forests in the abundance of pioneer, rare or economic species. These differ in detail, but ‘hysteresis’ – the forest memory – and other factors related to the concept of refugia apply to all these aspects of forest quality. Major hotspot refugia are crucial to the national framework of biodiversity, but local refugia, between the size of individual plants and single forest blocks, are crucial to local regeneration and sustainable use, as they shape the probability cloud which defines the anatomy of and processes within each species' range. Short-term sustainable use depends on local refugia; longer-term sustainability requires maintenance of refugia on a wider range of scale.

The implications of these phenomena to forest management are discussed in conclusion. Forest health is a multi-scale, but particularly a broad-scale, phenomenon. Local processes like the regeneration of forest under canopy gaps, are subordinate to larger-scale patterns and not determined simply by a match between species physiology and gap dynamics or patterns in the physical environment. Success of a species in a certain landscape does not automatically imply the species can be successful in similar conditions in a different landscape elsewhere: the context of the landscape in terms of the broader mosaic is also important. Managers, whether of plantations or natural forest, need to monitor, plan, and protect indigenous species on all scales. Forest managers need also to be aware of and work with the ‘forest memory’ factor. Protective measures for rare or economically threatened species should be based on current refugia and, like them, be arranged on all scales from single trees to large forest blocks.

Researchers need to pay more attention to processes between the ecological and biogeographical, if they are to provide information for managers which has a useful synergy with existing types of data. Exploration is needed of the anatomy of the ‘probability clouds’ defining the statistics of dispersal and regeneration of rare or threatened species with respect to parent populations. What are the chances of a mahogany establishing at a point 500 metres from a mother tree? How is this statistic influenced by soil type? There is much to be learnt on scales between the canopy and the Dahomey Gap.

Diameter frequency distributions (dfds) of 197 tree species in Korup National Park and 118 tree species in Douala–Edea Forest Reserve (based on c. 41 000 and 25 000 trees ≥ 30 cm gbh in 135 and 104 plots along four transects at each site, respectively) were individually classified into four groups of species according to their maximum diameters (≥99% trees <50, <70 and <90cm dbh, and > 1% ≥90cm dbh), and separated into two abundance classes (n≥50 and n = 20–49 (or 15–49). A fifth group, from those species in 3 and 4, was defined by a relative lack of small stems and given detailed attention. Dfds were evaluated for all species together on a transect and site basis, and compared for individual species across transects, and between sites for those group 5 species in common.

Analysis focussed on three large ectomycorrhizal species in groups 3 and 4, Microberlinia bisulcata, Tetraberlinia bifoliolata and T. moreliana (Caesalpiniaceae), which were restricted to, and dominated, the forest on the nutrient-poor soils of transect P at Korup. M. bisulcata, the most abundant, was markedly group 5 in character, with very few small trees, whilst the Tetraberlinia species had strong replacement potential. Although there were 27 species (42% of groups 3 & 4) at Korup and 19 (48%) at Douala–Edea in group 5, M. bisulcata was unique in its combination of large size, dfd, mycorrhizal habit, spatial distribution and soil association. The potential replacement of this species was contrasted also with a non-ectomycorrhizal caesalp with similar dfd that occurred on nutrient-richer soils, Erythrophleum ivorense.

Knowledge of the Caesalpiniaceae-rich la forêt biafriéene of Letouzey (which includes Korup and Douala–Edea) is reviewed, and the distribution, ecology and regeneration of the caesalps discussed. Structure and inferred dynamics of the forest at Korup are discussed with special regard to the ectomycorrhizal species.

All plants expend energy and nutrients in the production of secondary metabolites. Ecosystem-level studies on the production of metabolites in African rain forests are reviewed and factors influencing the production and distribution of phenolic compounds discussed. These compounds appear to influence feeding behaviour in primate herbivores and may also be important in mediating the rate of leaf litter decomposition and nutrient recycling. Many metabolites are biologically active and some have proved to have medicinal value: research on some plant families that occur widely in West African forests and which have proved to be particularly rich sources of new metabolites is reviewed.

The species composition, and the number of species of mammals, in the Guinea–Congo rain forests varies with locality. The highest species richness (up to 130 species) is found in parts of the West Central and East Central rain forest regions, and the lowest richness occurs in the Western Region and on the periphery of the rain forest zone. Rodents and bats contribute the greatest number of species (about 25% each), with lesser numbers of primates, duikers, and small insectivores; the relative numbers within each order is similar in all localities regardless of the composition and total numbers of species. The geographical distribution of each species is due to the alternating periods of forest fragmentation and expansion in the past, and the species-specific response to these environmental changes. There are about equal numbers of arboreal (30%), terrestrial (35%) and aerial (27%) species, each species within these categories having very specific habitat requirements within the forest. Most rain forest mammals are either frugivores (36%) or insectivores (37%): monkeys are mainly arboreal frugivores and folivores, and most prosimians are arboreal insectivores; forest duikers are terrestrial frugivores, and terrestrial rodents are frugivores or omnivores. Leaves, in spite of their abundance, are rarely eaten because of their low nutritive value and the difficulties of digesting them and consequently folivores (9%) are uncommon. Insects form the principal food for insect-bats and shrews. Mammals exhibit well-defined food partitioning in relation to food selection and body size, and therefore competition for food is probably minimal except during the dry season. Reproduction is more or less continuous in rodents and artiodactyis (with peaks of births at selected seasons); in contrast, monkeys and bats have fairly precise periods when births occur, which vary according to the species. The timing of reproductive events is regulated by the rainfall regime, availabiiity of food, and phyiogenetic characteristics such as the length of gestation and body size. Mammals interact with other rainforest organisms, principally as primary and secondary consumers, pollinators, and dispersers of fruits and seeds. Most species of rain forest mammals are solitary, but a few live in monospecific groups or in polyspecific groups. Forest disturbance in recent years (especially forest clearance and hunting) has had very detrimental effects on the geographical distribution and numbers of individuals of all species of forest mammals.

Ichthyogeography is the section of biogeography which seeks to interpret fish biodiversity in terms of present and past distributions and abundance. It can also have practical applications in fisheries management and conservation. There are thought to be more than one thousand species of fish in the 40 or so major rivers of the Guinea–Congo rain forest region, with correspondingly high levels of endemicity. In the Congo or Zaire river basin alone, the proportion of endemic fish species may exceed 80% (from a total of >690), but many more taxonomic and distributional data remain to be gathered.

Unlike terrestrial vertebrates, the diverse freshwater fishes of Guinea–Congo evidently have distributions which are closely confined by hydrography. Hence, fish distributions may be of particular help in corroborating or refuting postulated geological events, patterns and processes and in explaining associated aspects of rainforest evolution. From the 19th century until now, the ichthyogeography of Guinea–Congo has been diagnosed largely in terms of presumed post-Miocene geological sequences of fish taxa and their past and recent dispersal in relation to particular hydrological conditions: mainly riverine volume discharge, salinity and temperature. From this, the fish fauna is, by convention, divided into ‘provinces’ established on endemism, palaeogeography and supposed physical or ecological barriers to dispersal. However, in this paper it is argued that the traditional ichthyogeographical accounts which highlight endemism and dispersal are generally flawed. It is argued here that historical patterns of fish distribution can only be fully understood if phyletic (cladistic) data are taken into consideration. While Upper and Lower Guinea and the Zaïre basin may be defined in part on the basis of endemism there is a lack of taxonomic and distributional evidence to show that Guinea–Congo is itself a cohesive ichthyogeographical unit.

There is clearly a need for comparisons with fish distributions outside the rain forest zone of Guinea–Congo. African inter-provincial, trans-continental and inter-continental comparisons reveal distribution patterns which may relate more to pre-Miocene rather than post-Miocene geology or present-day rain forest ecology. Continental drift, notably between Africa and South America, probably led to the separation 85 million years ago of previously united fish populations. This may account for recent higher-level phyletic correspondences between the separate rain forest fish faunas of Africa and the neotropics.

Last, the so-called ‘marine intrusive’ fishes – which are normally excluded from zoogeographical consideration – merit a careful re-evaluation. While they may be regarded as an inconvenience in developing scientific hypotheses, such intrusives can comprise a remarkable 30% or more of the riverine fish fauna in Guinea–Congo. It seems that the widely accepted ecological divisions between marine and ‘primary freshwater’ fishes are not as clear-cut as has been supposed. In addition, zoogeographically critical marine, trans-Atlantic, phyletic relationships apparently exist. These are probably best interpreted by using area cladograms in the context of ocean basin development, rather than by referring solely to marine fish dispersal and the traditional continental and provincial ichthyogeography of Guinea–Congo.

The occurrence of ectomycorrhizal fungi in West African rain forest systems is demonstrated and linked to the presence of a range of caesalpinoid legumes and to members of the genus Uapaca. The vast array of boletes, members of the Russulaceae and Amanitaceae, and chanterelles is described. The virtual absence of ectomycorrhizal Cortinariaceae is emphasised. After an overview three separate yet interlinked studies are described: a survey of ectomycorrhizal fungi throughout the region (Thoen): ectomycorrhizal fungi of a single rain forest community (Watling) and the significance and relationships of the genus Russula in the Guinea–Congo Region (Buyck). Some comparisons are made with other rain forest areas of the world.

A window onto the vast array of saprotrophic fungi found in West Central Africa with particular reference to the Cameroon attempted in this paper. After an overview (Watling) three diverse groups are taken as examples of this array: xylariaceous fungi (Laessøe, Whalley & Watling); polypores (Ryvarden & Watling) and lepiotoid fungi (Watling). Some comparisons are made with other tropical forests. This approach contrasts a lignicolous group in the ascomycetes with one in the basidiomycetes and contrasts terricolous lepiotoid agarics with polyporaceous terricolous and lignicolous fungi.

This chapter has two linked objectives. The first is to sketch the process of human occupance of the lowland rain forest zone in West Africa, and illustrate (by means of a case study) the current circumstances of human settlement at the forest margins. The second is to bring the category ‘indigenous peoples’ under review, and to point out some of the pitfalls associated with the use of this term in the context of the West African forest frontier. All parts of the forest zone in West Africa have a long and complex history of occupation and use. In consequence, I shall argue that it is rarely, if ever, realistic to think of forest country as a residuum of empty land not yet passed into human ownership and use, but that it is equally distorting to treat the local groups currently found in possession of the forest edge as sole custodians. Forest resources have over time become part of the fabric of wider community life in West Africa, up to and beyond the national level, and narrow definitions of the category ‘indigenous peoples’ are to be avoided. Rather than search for putative forest indigenes, living in greater harmony with nature than incoming settlers (a usage reflecting Amazonian experience) it would be better, I suggest, to consider the term ‘indigenous peoples’ as covering all groups in West Africa with effective local knowledge of the forest. An eventual aim should be to understand how the experience of forest occupance has passed into the make-up of the institutions and values of civil society throughout the forest zone in West Africa today, regardless of whether or not these groups have a current direct connection to surviving areas of high forest. This is a much broader sociological agenda than conservationists have been prepared to tackle so far, but the threatened forests of West Africa are also complex human landscapes of considerable historical depth, and this human complexity is, in my judgement, no less worthy of understanding and conservation than the envelope of non-human biodiversity from which is has emerged. In fact, the two sets of issues are inextricably intertwined.