688
Climatic Geomorphology
rainforest can be replaced by the savanna and vice versa, and thus, the morphogenesic
changes (Tricart, 1974b). In the lower Congo and in the southern Ivory Coast, the forest
trends toward savanna (Tricart, 1974b). The opposite effect can be observed in Senegal.
On the basis of a review of the forest and savanna zones, Aubreville (1962) found
anomalous ecologic evidence in African humid tropics, as well as in the Guyana. These
anomalies are interpreted by this author as a lack of recuperation of the forest, after the
drier conditions occurred in the last glacial period. He pointed out that forest regions
drifted toward the south (the principal core in Angola), with some forest traces remaining
in the coastal areas of Guinea Gulf and Congo. This interpretation contradicts the idea of a
stable forest. A similar study was carried out by Butzer (1978), in which, during the glacial
optimum (18,000 BP) a large reduction of the extension of tropical forest occurred,
resulting in three bastions: Guinea-Sierra Leona-Liberia, Nigeria-Cameroon, and Zaire.
The work on the distribution of flora and fauna in the forest areas by Hamilton (1976),
provided important data about relevant variations of the central Africa forest in the last
20,000 year (Figure 25.42).
In the South American forest there are curious systems of speciation probably due to
changes in the extension of the forest in the upper Pleistocene era. The forest has been
fragmented into isolated sections in which speciation occurred (Figure 25.43) (Goudie,
1983a; Bradley, 1985; Crowley and North, 1991). In Brazil there is paleogeographic
evidence that supports the existence of xeromorphic plants in cerrados and tabuleiros in
central Brazil and in the western Amazon. The Barreiras Formation, consisting of
quartziferous sand and kaolin, which crops out in the area, was intensively incised during
the last glacial period as a result of the lowering of the base level resulting from the marine
regression. This erosion was favoured by the existence of a dry climate that produced the
disappearance of the forest in the lowlands along the Amazon River. This implies a
reduction of the area of forest during dry stages; there are forest refuges only in high areas
of the Guyana and Brazilian shield, where precipitation is higher. In the Holocene, climate
turned more humid and the Flandrian transgression raised the base level (Dias de AvilaPires, 1974; Tricart, 1985).
2.3. Geomorphologic evidences
2.3.1. Regolith and relict weathering profiles
The greatest weathering depths occur in the humid tropics and, normally, these profiles
were developed in the past, although actual climate conditions in the tropics can increase
weathering effects (Ollier and Pain, 1996). Some authors point out that the regolith
thickness in high areas is much smaller than in areas of gentle relief, while others, to the
contrary, indicate that in some cases the greatest depths are located in high areas, and in
the bottom of the valleys the bedrock crops out, because of the fluvial incision (Pain and
Ollier, 1981). From this, it can be deduced that the weathering depth depends on the
weathering intensity and erosion. On the other hand, very thick profiles need deep
groundwater to maintain an active weathering front (Ollier, 1988a,b).
In many known cases, weathering has been developed during the Tertiary and
Cretaceous periods, reaching thicknesses higher than 100 m. Most of these profiles are
Climatic Geomorphology
rainforest can be replaced by the savanna and vice versa, and thus, the morphogenesic
changes (Tricart, 1974b). In the lower Congo and in the southern Ivory Coast, the forest
trends toward savanna (Tricart, 1974b). The opposite effect can be observed in Senegal.
On the basis of a review of the forest and savanna zones, Aubreville (1962) found
anomalous ecologic evidence in African humid tropics, as well as in the Guyana. These
anomalies are interpreted by this author as a lack of recuperation of the forest, after the
drier conditions occurred in the last glacial period. He pointed out that forest regions
drifted toward the south (the principal core in Angola), with some forest traces remaining
in the coastal areas of Guinea Gulf and Congo. This interpretation contradicts the idea of a
stable forest. A similar study was carried out by Butzer (1978), in which, during the glacial
optimum (18,000 BP) a large reduction of the extension of tropical forest occurred,
resulting in three bastions: Guinea-Sierra Leona-Liberia, Nigeria-Cameroon, and Zaire.
The work on the distribution of flora and fauna in the forest areas by Hamilton (1976),
provided important data about relevant variations of the central Africa forest in the last
20,000 year (Figure 25.42).
In the South American forest there are curious systems of speciation probably due to
changes in the extension of the forest in the upper Pleistocene era. The forest has been
fragmented into isolated sections in which speciation occurred (Figure 25.43) (Goudie,
1983a; Bradley, 1985; Crowley and North, 1991). In Brazil there is paleogeographic
evidence that supports the existence of xeromorphic plants in cerrados and tabuleiros in
central Brazil and in the western Amazon. The Barreiras Formation, consisting of
quartziferous sand and kaolin, which crops out in the area, was intensively incised during
the last glacial period as a result of the lowering of the base level resulting from the marine
regression. This erosion was favoured by the existence of a dry climate that produced the
disappearance of the forest in the lowlands along the Amazon River. This implies a
reduction of the area of forest during dry stages; there are forest refuges only in high areas
of the Guyana and Brazilian shield, where precipitation is higher. In the Holocene, climate
turned more humid and the Flandrian transgression raised the base level (Dias de AvilaPires, 1974; Tricart, 1985).
2.3. Geomorphologic evidences
2.3.1. Regolith and relict weathering profiles
The greatest weathering depths occur in the humid tropics and, normally, these profiles
were developed in the past, although actual climate conditions in the tropics can increase
weathering effects (Ollier and Pain, 1996). Some authors point out that the regolith
thickness in high areas is much smaller than in areas of gentle relief, while others, to the
contrary, indicate that in some cases the greatest depths are located in high areas, and in
the bottom of the valleys the bedrock crops out, because of the fluvial incision (Pain and
Ollier, 1981). From this, it can be deduced that the weathering depth depends on the
weathering intensity and erosion. On the other hand, very thick profiles need deep
groundwater to maintain an active weathering front (Ollier, 1988a,b).
In many known cases, weathering has been developed during the Tertiary and
Cretaceous periods, reaching thicknesses higher than 100 m. Most of these profiles are