-- !--i Paleoclimatology and

Paleometeorology : Modern and Past Patterns of Global Atmospheric Transport edited by

d-. . I I ' e : . Margaret Leinen Graduate School of Oceanography, University of Rhode Island, Narragansett, RI; U.S.A. ~

and

Michael Sarnthein Geologisches Paleontologisches Institut, Christian Albrechts Universität, Kiel, F.R.G.

NATO, AsSI Series Advanced Science Institutes Series

Series C Mathematical and Physical Sciences - Vol. 282 (1 989) a,7: + $' , .L. : . . s .

I /'L Fonds Documentaire, .-

, .

Kluwer Academic Publishers

Dbrdrecht I Boston I London . * ,>c . ;? E - . i

Published in cooperation with NATO Scientific Affairs Division . ,

.

LATE QUATERNARY CLIMATIC CHANGES IN THE AFRICAN RAIN FOREST : FOREST REFUGIA AND THE MAJOR ROLE OF SEA SURFACE TEMPERATURE VAFlIATIONS.

O. R. S.T. O. Ri. Foiids Documentaire

Jean MALEY ORSTOM, UR A3 et CNRS, UA 327 Laboratoire de Palynologie Universite des Sciences et Techniques du Languedoc 34060 MONTPELLIER cedex, FRANCE.

ABSTRACT. The question of forest refugia in Africa is briefly presented from a historica'l and evolutionary point of view. Main pollen evidence, which goes back to about 25-28,000 yr BP, is presented for two lakes situated in African lowland rain forest. - The pollen data from Lake Bosumtwi (Ghana) show the disappearance of rain forest from ca. 28,000 to ca. 9000 yr BP. During this time interval the vegetation was a grassland of the montane type with' sparse clumps of trees. The're is synchronism between montane vegetation disappearance and rain forest reappearance. This phenomenon occurred abruptly between about 9000 and 8500 yr BP. - In Lake Barombi Mbo (West Cameroon) the pollen data show clearly that from ca. 24,000 yr BP until recent time, rain forest persisted with limited variations, and thus, this area represents a refuge area. One also -notes an extension of montane vegetation to l o w elevation which disappears near the beginning of Holocene time. After a short synthesis of the palaeobiogeography of Afromontane vegetation, one discusses the palaeoclimatology. In the light of the present-day annual climate anomalies which relate mainly to sea surface temperature variations and upwelling of cold water in the Guinea Gulf, one points out the drying and cooling role of stratiform cloud cover, particularly during the time of fragpentation of the forest area. For the Ghanean rain forest one also discusses on the weather types responsible for the forest extension after 9000 yr BP and during the ma.jor transgression of BosUmtwi lake during mid-Holocene time. One relates these two phenomena to the disappearance of the present-day summer "little dry season". The reappearance of this "little dry season" in late Holocene time, linked to the deterioration of the water budget, is probably responsible for the abrupt lacustrine regressions and also for the opening of the "Dahomey Gap".

1. INTRODUCTION. THE QUESTION OF FOREST REFUGIA

In the equatorial zone, the rain forest biome has long been considered

585

M: L e i m and M. Sumthein (eh.), Paleoclimatology and Pakmteorology: M&rn and Past Patterns of Global Atmospheric Transport* 585-616. 8 1989 by K h e r Academic Publishers.

586

t o be a quasi-f ixed e n t i t y , s h e l t e r e d from c l i m a t i c f l u c t u a t i o n s . In Afr ica , b o t a n i s t s such a s AUBREVILLE (1949,1962) o r SCHNELL (1950) and z o o l o g i s t s a s BOOTH (1958), MOREAU (1963,1966) o r CARCASSON ( 1 9 6 4 ) who d e s c r i b e d t h e l a r g e f l o r i s t i c and f a u n i s t i c h e t e r o g e n e i t y of t h i s region, w e r e among t h e f i r s t who came t o t h e fol lowing conclusion : t h e f o r e s t biome had a l s o undergone profound m o d i f i c a t i o n s . I n t h e end, t h e c o n t i n u a t i o n of r e sea rch led t o t h e hypo thes i s t h a t during c l i m a t i c a l l y un favorab le p e r i o d s , t h e b i o l o g i c a l r i c h n e s s of t h e f o r e s t was conserved i n r e fug ia , p r i v i l e g e d s e c t o r s where t h e cl imate would have remained f a v o r a b l e . The f i r s t schemat i c maps of f o r e s t r e fug ia i n e q u a t o r i a l Afr ica were publ ished by AUBREVILLE (1949, p.66; 1962, p .62 ) .

Neve r the l e s s , u n t i l r e c e n t l y , r e s e a r c h on t h e q u e s t i o n of f o r e s t r e f u g i a have inc reased , e s p e c i a l l y i n South America, i n t he Amazon F o r e s t (PRANCE, 1982; WHITMORE et PRANCE,1987), f i r s t l y i n o r d e r t o l o c a t e t h e p o s i t i o n of r e f u g i a , s econd ly t o s t u d y t h e s p e c i a t i o n p rocesses a s s o c i a t e d w i t h t h e geograph ica l i s o l a t i o n of r e f u g i a . Indeed, f o r many au tho r s , t h e b i o l o g i c r i c h n e s s of f o r e s t biome cou ld a t l e a s t p a r t l y be explained u s i n g t h e model of a l l o p a t r i c s p e c i a t i o n ( v i c a r i a n c e ) (HAFFER, 1982) . However, a c c o r d i n g t o t h e r e s u l t s of s t u d i e s of animal and p l a n t groups, t h e e x i s t e n c e of r e f u g i a has been quest ioned because t h e proposed a r e a s were no t always superposed o r had s u r f a c e s of d i f f e r e n t s l z e (cf .examples g iven i n WHITMORE e t PRANCE,1987). Y e t t h e s e a u t h o r s c o n s i d e r t h a t t h o s e d i f f e r e n c e s cou ld be l i n k e d mainly t o t h e dynamics of each animal o r p l a n t group, without denying t h e r e f u g i a . Moreover, some au tho r s , a s ENDLER (19821, cons ide r ing t h e Afr ican f o r e s t model, e s t i m a t e t h a t i t s b i o l o g i c a l r i chness could r e s u l t uniquely from p a r a p a t r i c s p e c i a t i o n s . Indeed, t h e necessary i s o l a t i o n f o r t h e emergence of new s p e c i e s could be achieved only by t h e presence i n t h i s biome of numerous eco log ica l n i c h e s (FEDOROV, 1 9 6 6 ; RICHARDS, 1969; e t c ) . I n r e f u t i n g some of ENDLER's arguments (1982) , MAYR and O'HARA (1986) re-aff i rmed evidence suppor t ing a l l o p a t r i c s p e c i a t i o n due t o f r agmen ta t ion of t h e African r a i n f o r e s t a r e a .

A l l t h e s e a u t h o r s f i n a l l y e s t i m a t e t h a t t h e b e t t e r arguments suppor t ing t h e former e x i s t e n c e of r e f u g i a and f r agmen ta t ion of t h e I a i n f o r e s t a r ea could be provided by po l l en ana lyses on d e p o s i t s from a r i d phases.

The f i r s t aim of t h i s paper i s t o p r e s e n t t h e main po l l en r e s u l t s answering t h i s q u e s t i o n . The second aim i s t o d i s c u s s the c l i m a t i c c o n d i t i o n s which l e d t o f r agmen ta t ion of t h e Af r i can r a i n f o r e s t i n t o i s o l a t e d r e fug ia and l a t e r on t o f o r e s t r eco lon iza t ion .

2. POLLEN ANALYSES I N AFRICAN R A I N FOREST

During t h e l a s t few yea r s , p o l l e n ana lyses w e r e c a r r i e d o u t on upper Qua te rna ry d e p o s i t s from two l a k e s s i t u a t e d a t low e l e v a t i o n i n the lowland Af r i can r a i n f o r e s t : t h e Bosumtwi l a k e i n Ghana and t h e Barombi Mbo l a k e i n W e s t Cameroon ( F i g . 1 ) .

587 * .

F i g u r e 1 - Diagram of lowland r a i n f o r e s t r e f u g i a i n e q u a t o r i a l Af r i ca d u r i n g maximum of t h e l a s t g r e a t d r y and c o l d p e r i o d (about 18,000 y r B P ) . The r e f u g i a of Upper Guinea and E a s t e r n - , Za ï r e a r e adapted from ENDLER (1982) and MAYR et O'HARA (1986), from d a t a of MOREAU (1966 ,1969) and HALL et MOREAU (1970). The modern cond i t ions ( f o r e s t boundary, i nc luded savannas and main i s o h y e t s ) a r e adapted from WHITE (1983, f i g . 5 and e x t r a c t e d from MALEY, 1987) .

map) ( f i g u r e 1 :

2.1. Lake Bosumtwi in Ghana

This lake with a water level near lOOm ASL (Above Sea Level) is situated in semi-deciduous rain forest that is characterized by the families of Ulmaceae, particularly with Celtis, and Sterculiaceae with Triplochiton and other genera (HALL et SWAINE, 1981) . Climatically, this, type of forest is adapted to a slightly less humid conditions than the evergreen type which will be presented below ( 5 2.2). The semi-deciduous forest appears when the yearly number of "dry" months (rainfall below about 100") reach three. The two months of the "little dry season" in August and September (5 4.2; Fig.9) are quasi rainless, but because the atmospheric humidity remains high, these two months must not be added (see the Gabon forest example, § 4.2) to the three months dry season of northern winter (December to February).

The core studied from lake Bosumtwi, has a length of about 17m and reaches back to about 28,000 yr BP (Fig.2). The geological study of this corer and outcrops of lacustrine deposits, permitted the reconstruction of the lacustrine fluctuations which were very distinct, with a maximum lake level in middle Holocene time and a very low lake level from about 20,000 to 15,000 yr BP (TALBOT et a1.,1984; TALBOT et KELTS, 1986; MALEY, 1987) .

The principal pollen results (MALEY et LIVINGSTONE,1983; TALBOT et al., 1984; MALEY,1987) (Fig.2 and 3) clearly show that before about 9000 yr BPI forest was not present in this region. Indeed, between the present and about 8500 yr BP arboreal pollen percentages oscillated from 75 to ES%, while before 9000 yr BP arboreal pollen percentages wereagenerally below or close to 25%, except near the base of the core, where they were close to 50%. During the period between about 19,000 and 15,000 yr BPI arboreal pollen percentages reached minimum values of about 4 and 5%. At the same time trees had almost completely disappeared from the landscape and had been replaced by the herbaceous plants,-essentially Gramineae and Cyperaceae which reached frequencies of 91 to 94%. Today such percentages of pollen grains from herbaceous and arboreal plants are common in the Sahelian zone (MALEY,1981).

Other important evidence resulting from pollen analyses is the presence of a montane element in the period from the core base to its quasi disappearance around 8500 yr BPI near the time when the rain forest reappeared (Fig. 3) . This montane element includes the mountain olive, Olea hochstetteri, which today are only present 700 km westward, near 1200m ASL on the Momy Mountain in the Dan Massif of western Ivory Coast (SCHNELL, 1977) . Eastward, the nearest modern lbcatfon of this tree is in Nigeria on the Jos Plateau, and further eastward mainly in the mountains of the Cameroonian Ridge (LETOUZEY, 1968). Until about 8500 yr BPI the presence of a montane element at low elevation around the Bosumtwi lake, where the highest surrounding hills reach maximally elevations from 400 to 550m ASL, imply a lowering- of the .montane. vegetation belt with minimally 600m. This corresponds to a temperature decrease of 3O to 4OC (MALEY et LIVINGSTONE, 1983) when applying a mean temperature gradient of O. 6OC fo r every loom displacement of the vegetation belt (WALTER et BRECKLE, 1985) (5 4.1). This temperature decrease has also been reached by .I _. ' k J L i

Y E A R S

1 --i

A

~.

8.P. (L 1000

0-

5-

10 -

15-

20-

25 -

I LAK E-LWEL / VARIATIONS I

B

ILLEN and SPORES-

.L.

:

589

Figure 2 - Lake Bosumtwi in Ghana. A - Stratigraphy of corè B-7, from TALBOT et a1.(1984). B - Lake level variations adapted from TALBOT et KELTS (1987); before about 15,000 yr BP the curve is modified by reference to pollen data (cf .Fig.3) . C - Pollen data from TALBOT et al. (1984) and MALEY (1987). P -

Cuticles of Pooideae (Afro-montane Gramineae) by PALMER, in TALBOT et al. (1984).

590

LAKE BOSUMTUI POLLEN ANALYSES (J.MALEY)

Figure 3 - Lake Bosumtwi pollen analyses (J.MALEY). The chronology of the core is based on 15 radiocarbon dates (after TALBOT et a l . , 1984).

,,

59 1

PALMER (1982; in TALBOT et a1.,1984) by determining in the same section of the core the grass cuticle fragments belonging to the tribe of Pooideae. Today this tribe occurs in the tropical zone only on the high mountains (CLAYTON, 1976; LIVINGSTONE et CLAYTON, 1980) , such as on Mount Cameroon where representatives have been recorded only above ca. 2000m altitude (LETOUZEY, 1968,1985). The presence of Pooideae flora below 550m altitude could involve an important decrease of temperature (TALBOT et al. , 1984) which should be approximately 6OC, a value comparable to estimates from East Africa ( E 3).

Thus, before 9000 yr BP the forest was absent and it gave way to a mountain type grassland with sparse clumps of trees. Near the base of the core, dated about 27 - 28,000 yr BPI arboreal pollen percentages of about 50% are indicative of the presence of some kind of mountain forest. A remarkable feature is that the scattered trees in the open vegetation belonged in great majority to the semi- deciduous forest flora and not to the Sudano-Guinean savanna flora. Under modern conditions, only the Guinean montane grasslands of middle elevation (submontane belt) include isolated clumps of trees which, in addition to typical mountain species, contain a large set of species which are found from low elevations to the submontane belt (SCHNELL, 1977; LETOUZEY, 1968, 1985) . 2.2. Lake Barombi Mbo in West Cameroon

Lake Barombi Mbo has a water level near 300m ASL and is situated in the lowland rain forest of West Cameroon (Fig.1). Around the lake two principal forest formations can be recognized in an area with about 10 km radius (RICHARDS, 1963; LETOUZEY, 1968, 1985; D.THOMAS, pers. commun. ) :

Guineo - Congolan type. It spreads around the Bay of Biafra from east Nigeria to as far south as the boundary of Cameroon and Equatorial Guinea (LETOUZEY, 1985; WHITE, 1983). This forest type is characterized by its richness in Leguminosae, especially Caesalpi- niaceae and is present when the dry season does not exceed two months. - ous fo- (5 2.1) forms islets in the evergreen

S o far only the lower half of the longest core (BM-6: 23.50m, from the base to about 11,000 yr BP) has been studied in detail (28 samples spaced every 30 or 60 cm have been analyzed : BRENAC,1988; MALEY et al.,1988(b), to be published). Samples at 2 or 3m intervals have also been studied from the upper part of the core (MALEY et BRENAC, 1987). Below 21m depth in the core, perturbed section interrupts the continuity of pollen spectra . However the dates obtained near the base and the pollen content (Fig.4) show that the lowermost part of the core is probably reversed (MALEY et a1.,1988(b) to be published).

The following principal results are to be noted (Fig.4): z!axLL From the base of the core (greater than 24,000 yr BP) to a level with an age of about 20,000 yr BPI pollen grains of Gramineae, Cyperaceae

- eve- forest of B e e a n tvDe is a richer variety of the

forest which remains largely dominant (LETOUZEY,1985). - I I

-. -

I , ' , '* t ' . . j . 'i-

- 592 593

Lake Barombi Mbo (West-Cameroon) Pollen Analyses o f Late Pleistocene

- . L.A- . - -

o 207. o IO 201 O I O Z O IO zo33z O 20 40 60 W.

Zone I I I

Zone I I b

Zone I I a

Zone I

Figure 4- -. Lake Barombi Mbo p o l l e n ana lyses (P.BRENAC) . (from BRENAC, 1988; MALEY e t a l . , I988 ,b I t o be pub l i shed) .

and s h o r e l i n e p l a n t s a r e sca rce . Gramineae p o l l e n i s r ep resen ted from 6- t o 15%. Among t h e a r b o r e a l p o l l e n , t h e moutain o l i v e , O l e a h o c h s t e t t e r i , is abundant with percentages between 1 0 and 30%. Phoenix r e c l i n a t a (Palmae), which i s found today i n t h e Cameroon mountains f r e q u e n t l y a s s o c i a t e d with Olea h o c h s t e t t e r i (LETOUZEY, 1978), is a l s o r e l a t i v e l y f r equen t i n t h i s s e c t i o n . Besides t h e s e two montane taxa, p o l l e n of f o r e s t t a x a is abundant with pe rcen tages from 60 t o 75%. Among these, Caesalpiniaceae and Euphorbiaceae a r e p a r t i c u l a r l y w e l l represented (BRENAC, 1988) .

Under modern cond i t ions , t h e a s s o c i a t i o n of moutain t a x a i n combination w i t h t y p i c a l lowland evergreen f o r e s t f l o r a i s found i n Cameroon, f o r example, on t h e h i l l s ' summits around Yaounde (between 900 and 1200m ASL) (ACHOUNDONG, 1 9 8 5 ) . On t h e p l a i n around t h e s e h i l l s t h e f o r e s t i s semi-deciduous, bu t on t h e h i l l s ' summits c l i m a t e i s more humid and cool because of t h e high frequency of c louds which favor t h i s p a r t i c u l a r c loud f o r e s t a s s o c i a t i o n (BRENAC, 1988) . zsmLLL Around 20,000 y r BP, a very sha rp change occurs i n t h e p o l l e n s p e c t r a . Percentages of mountain o l i v e p o l l e n g r a i n s decrease t o va lues of 5 t o 15%. Fores t t axa dec rease a s w e l l , p a r t i c u l a r l y Caesa lp in i aceae and Euphorbiaceae. Percentages of a r b o r e a l p o l l e n g r a i n s o s c i l l a t e around 4 0 % u n t i l ca . 1 4 , 0 0 0 y r BP. During t h e same pe r iod , g r a s s p o l l e n g r a i n s i n c r e a s e t o about 25%. It i s t h e same f o r Cyperaceae and o the r a q u a t i c p l a n t s .

These d a t a a r e i n d i c a t i o n s of a r e l a t i v e l y d r y and cool c l ima te . The i n c r e a s e of Cyperaceae and s h o r e l i n e p l a n t s i s probably r e l a t e d t o a lower water l e v e l , which i s i n accordance with a decrease i n r a i n f a l l . By analogy with p o l l e n a n a l y s e s of modern samples of s o i l s and l a c u s t r i n e sed imen t s (BRENAC, 1988) , t h e i n c r e a s e i n herbaceous p l a n t s i s i n d i c a t i v e of t h e development of more open f o r e s t vege ta t ion o r t h e development of a mosaic p a t t e r n of f o r e s t and open formations. - Between c a . 1 4 , 0 0 0 and 1 0 , 0 0 0 y r BP t h e f o r e s t t a x a i n c r e a s e again, w i t h p e r c e n t a g e s c l o s e t o 6 0 % . Among t h e t r e e t a x a , some r e p r e s e n t a t i v e s of t h e semi-deciduous f o r e s t , such a s C e l t i s , r e p l a c e those of t h e evergreen f o r e s t p re sen t b e f o r e 20,000 y r BP. One notes a l s o p i o n e e r f o r e s t t a x a t y p i c a l of secondary n a t u r a l fo rma t ions . Around t h e beginning of Holocene t i m e , Gramineae percentages decrease t o abou t 20%. I n Zone III, p e r c e n t a g e s of mountain o l i v e p o l l e n i n c r e a s e a g a i n around 13,000 y r BP t o d i s a p p e a r a g a i n n e a r t h e beginning of Holocene (BRENAC, p r i v . commun. ) . One can conclude t h a t t h i s pe r iod was humid again, bu t not a s c o o l a s be fo re 20,000 y r BP.

From about 1 0 , 0 0 0 y r BP t o t h e p r e s e n t , t h e few d a t a so f a r a v a i l a b l e show a r e l a t i v e l y s t a b l e v e g e t a t i o n , domina ted by semi-deciduous f o r e s t elements accompanied by many t a x a of secondary n a t u r a l formations (MALEY e t BRENAC, 1 9 8 7 ) .

2.3. Conclusions

T h e po l l en d a t a presented he re a r e p a r t i c u l a r l y important because they

594 1

595 * -

show that during the last great arid phase, from about 20,000 to 14,000 yr BP, the rain forest disappeared in Ghana, particularly in the Bosumtwi sector, but persisted in West Cameroon. These conclusions have been associated with various biogeographic data obtained from studies on plants, birds, mammals, amphibians, butterflies, etc., from Cameroon and other parts of the forested zone, as well as palaeoclimatic data, in order to establish a new schematic map of lowland forest refugia in equatorial Africa during the last great arid phase, about 18,000 yr BP (Fig.1) (see MALEY, 1987, for a more detailed discussion). Another important result is the spread to low elevation of a montane element that are characterized by Olea hochstetteri. This indicates a temperature cooling which seems similar f o r the two sites studied.

3. PALAEOBIOGEOGRAPHY OF AFROMONTANE VEGETATION

To understand the problem of extension of montane vegetation to lowland, the modern composition of the African montane flora and fauna of middle altitudes from about lOOOm to 3000m ASL must be recalled. Indeed, between the different mountain massifs of equatorial Africa, a great similarity has often been noted. For example, according to HALL (19731, the plant species common to Mount Cameroon and Eastern Africa are 53% for the montane forest and 49% for the montane grassland. In order to explain these similarities, many authors have assumed that during climatic changes and especially during cooler periods of the Quaternary, the montane floras and faunas must have extended to the lowlands which facilitated migration between mountain massifs (MOREAU, 1966;WHITE, 1981; MALEY, 1987) .

The extension of montane vegetation to lower elevation in Eastern Africa has been known for several decades, and for this reason temperatures are believed to have been 5' to 9OC lower (COETZEE,19'64; VAN ZINDEREN BAKKER et COETZEE, 1972; HAMILTON, 1973, 1902; FLENLEY, 1979) . Moreover, in southern Congo Republic, (CARATINI et GIRESSE,1979; ELENGA et a1.,1987) and in northern Angola (VAN ZINDEREN BAKKER et CLARK, 1962) , several pollen analyses also clearly show extensions of montane vegetations to low elevation (MALEY, 1987) . The most important data has been obtained on the Plateaux BatBk6, about 40 km north of Brazzaville, in a small depression near 600m ASL (ELENGA et a l . , 1987). The pollen analyses carried out on a short core show that from the base until around the beginning of Holocene time, the pollen spectra were dominated by Afromontane taxa ; the pollen grains of Podocarpus milanjianus (5yn.P. latifolius) , Ilex mitis and Olea welwitchii (syn. O.hochstetter2, cf .TROUPIN, 1985) constituted about 60% of the pollen spectra, dominated by Podocarpus with about 5O%(ELENGA et al.,1987).

Between the mountain massifs of Eastern Africa and Cameroon, several typical Afromontane taxa had been observed in various isolated stations (WHITE,1981), on the southern rim of the, Zaïre river catchment, and along the Guinea Gulf on the ridge of hills connecting Angola to Cameroon. For Podocarpus latifolius, there is a station near

700m ASL on the Congolese slope of Chaillu Mountain (MALEX et a1.1988 (a),to be published) and another about 900m ASL near the top of an isolated inselberg in south Cameroon, near the Gabon border (LETOUZEY, 1968) . The cumulative evidence from these stations and the pollen data presented above show the pattern of a preferential junction way which could have operated intermittently-during the Quaternary between East Africa and Angola, and subsequently expanded to the hill ridge of Mayombe, Chaillu, Monts de Cristal until eventually the Cameroon mountains (MALEY,1987; MALEY et a1.1988 (a) , to be published) (Fig.5).

4,PALAEOCLIMATOLOGY

4.1. Introduction. The hypothesis of lapse rate variation

During the Last Glacial Maximum (LGM) (about 18,000 yr BP) the extension of glaciers and vegetation belts to altitudes much lower than those of today are now widely accepted data (see for example the synthesis of FLENLEY, 1979). Many authors have tryed to explain the temperature lowering which can be deduced from these data (generally between So and 9OC) with palaeoclimatic models (WEBSTER et STRETEN, 1978; WHITMORE et al. , 1982; FLENLEY, 1984; RIND et PETEET, 1985; MORLEY et FLENLEY, 1987; etc). The paper of WEBSTER and STRETEN (1978) is a good example of one of these essays. The authors (op cit.) discuss about tropical Australasian palaeoclimates during the LGM, and like for tropical Africa, much of the data show a 6' to 8"C lowering of temperatures in New Guinea mountains. For the surrounding tropical oceans, the data from the C L I W Group (1976,1981) suggest only a 2OC cooling or less. The value is similar for the seas around tropical Africa, except in a small sector of the Guinea Gulf where a 3OC cooling is indicated. WEBSTER and STRETEN (op cit.) try to explain these opposite results from Australasia with several detailed palaeoclimatic reconstructions.

The principal hypothesis that is discussed in detail is the variation of the vertical lapse rate (WEBSTER et STRETEN, op cit.). About this physical phenomenon, WALTER and BRECKLE (1985, their Fig.120) used many temperature records from various mountain slopes in Venezuela to obtain a linear temperature elevation relationship, and from this relation calculated a vertical lapse rate of 0.57OC per lOOm from sea level to 5000m. Using the temperature, recoFded during vertical ascent in free atmosphere for several stations in Australia and New Guinea, WEBSTER and STRETEN (op cit.) have shown that this value (about 0.6'C/lOOm) corresponds to the moist adiabatic lapse rate, but that in dry atmosphere the lapse rate is steeper and close to 0.8°C/100m. The principal point is that this dry atmosphere lapse rate fits quite well the palaeoclimatic data for the LGM. Because the climate was dry at about 18,000 yr BP, WHITMORE et al. (1982) , FLENLEY (1984) and MORLEY and FLENLEY (1987), relying on this hypothesis, have estimated that the atmosphere was dry also at this time and, for this reason, had a steeper lapse rate. However WEBSTER and STRETEN (op cit)

596

- i

Figure 5 - Distribution of the nine African regional mountain systems. I, West-African; II, Cameroon-Jos; III, Ethiopian; IV, Imatongs-Kenya-Usambara; V, Ruwenzori-Kivu; VI, Uluguru-Mlanje; VII, Chimanimani; VIII, Drakensberg; IX, Angolan. The Stars show many distant populations of Afromontane plant species outside the large mountain systems represented by Mack areu (adapted and completed from WHITE, 1978, 1981). The two arrows schematically represent a possible migration path. of Afromontane taxa from East Africa to Angola and subsequently via'Cameroon.to West Africa.

591

criticize this hypothesis because ( o p cit. ,p -305) the thermodynamics determining the atmospheric structure in the vertical are governed by strict physical laws. The vertical temperature profile are expected to be close to moist adiabatic even if the atmosphere were drier, as indicated by the winter temperature-height diagrams of Darwin, Cloncurry and Charleville. WEBSTER and STRETEN (op cit.) also expect similarity between the temperatures above the adjacent ocean regions and at corresponding levels on elevated terrain.

Additionally, in an important paper, RIND and PETEET (1985) conclude (ibid.,p.l8) that at low latitudes the current lapse rate is consistently close to the moist adiabatic value and that, for this reason, the lapse rate would not have changed if sea-surface temperatures remained warm during the LGM.

Consequently an alternative hypothesis is necessary to explain the LGM data. The starting point of the alternative hypothesis presented here is that one observes today on some mountains a natural extension of mountain vegetation to lower elevations. The new hypothesis , which has been already formulated (MALEY et LIVINGSTONE, 1983; MALEY, 1984,1987) , is based partly on the fact that ecologists studying present-day mountain conditions have demonstrated the primordial role played by the cloud covers and fogs (GRUBB et WHITMORE, 1966; BAYNTON, 1968; GRUBB, 1971, 1974, 1977).

4.2. Some present-day models of localized montane extensions to lower elevations

The montane biotopes appear generally above lOOOm ASL; this limit forms a fundamental biological barrier which is characterized by changes in flora and fauna and also commonly by physionomical modifications (MOREAU, 1966; HOWARD, 1970; GRUBB, 1971,1974; WHITMORE, 1975; LEIGH, 1975; BERNARDI, 1979; etc.) . However, one can locally observe the extension of normally montane faunas or floras to low elevation.

Outside of Africa, there are many examples of this phenomenon in the Antilles islands and around the Caribbean Sea (BAYNTON,1968; HOWARD,1970; etc.). In the northern tip of Colombia, SUGDEN (1982) describes a cloud forest on the Serrania de Macuira which attains a maximum elevation of 865m about 25 km inland from the coast. The slopes from about 500m to the summit are covered by a mountain forest formation. In the lowland the mean monthly temperature is 28OC and varies little throughout the year (SUGDEN, op cit.). At 500m ASL SUGDEN (op cit.) recorded a mean temperature of about 22.5OC and so, one obtains a temperature lowering of 5.5OC for 500m. Usually, with a lapse rate of 0.6'C / lOOm (see above 5 4.1) one can calculate a temperature lowering of 3OC for an elevation of 500m. The extra 2.5OC of cooling must be related to the extensive cloud formation and fog which enshroud the summit.

Other examples exist on the Atlantic coast of Africa, but apparently, without temperature measurements. The hills of Freetown in Sierra Leone present one example. These hills rise from the sea to an elevation of about 900m ASL and, from 500m ASL upwards, they support a

598

set of montane p l a n t s (MORTON,1968), such a s Olea h o c h s t e t t e r i (HEPPER,1963). Fu r the r t o t h e e a s t on t h e seaward south f l a n k of Mount Cameroon, t y p i c a l montane trees (THOMAS, 1986) and birds (SERLE, 1 9 6 4 ) a r e found above 5 0 0 m ASL. F u r t h e r t o t h e s o u t h on c h e Angola Escarpment, which rises above t h e s e a t o t h e l e v e l of t h e P l a t e a u above 1000m, a c loud f o r e s t i s found from 200 or 300m ASL with t h e i m p l i c a t i o n t h a t montane c o n d i t i o n s a l r e a d y appea r a t t h i s low e l e v a t i o n (EXELL e t MENDONCA, 1 9 3 7 ; AIRY-SHAW, 1947; WHITE e t BERGER,1978). Other examples e x i s t i n E a s t A f r i c a n e a r t h e Ind ian Ocean on t h e Usambaras Mountains (MOREAU,1935, 1938, 1 9 6 6 ) . A l l t h e s e occurences show t h a t t h i s i s a non- fo r tu i tous e c o l o g i c a l phenomenon which al lows t h e ex tens ion of t h e c l i m a t i c c o n d i t i o n s normally found above 1 0 0 0 t o 1500m t o lower e l e v a t i o n s . The d a t a p re sen ted above f o r t h e S e r r a n i a de Macuira i n Colombia a r e a n example of magnitude of temperature reduct ion i n such environment.

Most a u t h o r s who have c o n s i d e r e d t h i s phenomenon have a t t r i b u t e d it t o p e r s i s t e n c y of c l o u d c o v e r and f o g which a r e p a r t i c u l a r l y f r e q u e n t on s l o p e s f a c i n g t h e s e a (AIRY-SHAWI1947 ; MOREAU, 1966 ; SERLE, 1 9 6 4 ; HOWARD,1970; GRUBB, 1971; etc) . The Angolan Escarpment i s a p a r t i c u l a r l y important model because it shows how t h e abundance of c l o u d cover depends d i r e c t l y on t h e low c louds coming from t h e s e a where t h e c o l d Benguela c u r r e n t f lows. HOFLICH (1972) has shown how t h i s c u r r e n t produces a t h i c k mantle of s t r a t i f o r m clouds which i n f l u e n c e g r e a t l y t h e c l i m a t e of t h e neighboring con t inen t by reducing t h e r a i n f a l l and lowering t h e temperature t h e r e .

. . 599

4.3. The p resen t day c l i m a t i c r o l e of upwelling and s t r a t i f o r m clouds

Recent r e sea rch i n t h e t r o p i c a l A t l a n t i c s e c t o r have shown t h a t t h e i n t e r a n n u a l v a r i a b i l i t y of t h e c l i m a t e i s r e l a t e d t o enhancement o r r e d u c t i o n of t h e a n n u a l c y c l e (HASTEYRATH, 1 9 8 4 ; NICHOLSON and ENTEKHABI,1987) . For example, HIRST and, HASTENRATH (1983a, b ) , s tudy ing t h e v a r i a t i o n of r a i n f a l l over t h e ZaPre r i v e r catchment, have found p o s i t i v e c o r r e l a t i o n s between t h e e v o l u t i o n of t h e two p r i n c i p a l r a i n y seasons and sea s u r f a c e cemperature (SST) o f f Angola. The pe r iods with warmer ( co lde r ) SST a r e c o r r e l a t e d with above (below) normal r a i n f a l l s . SST modulate r a i n f a l l by c o n t r o l l i n g moi s tu re and s t a b i l i t y i n t h e lower atmosphere which a r e a l s o r e l a t e d t o c loud iness . But ano the r p o s i t i v e c o r r e l a t i o n e x i s t between upper a i r cond i t ions and r a i n f a l l over t h e Za ï r e catchment (HIRST et HASTENRATH, 1983a) . Probably l a r g e s c a l e a tmospheric c b n d i t i o n s i n f l u e n c e both parameters (NICHOLSON e t ENTEKHABI, 1987) . I n t h i s way, SST v a r i a t i o n s and upwelling of c o l d waters (Fig.6) a r e indeed d i r e c t l y and remotely governed by t h e i n t e n s i t y of t h e winds which a r e r e l a t e d t o t h e a c t i v i t y of t h e South A t l a n t i c High (MERLE,1980; HASTENRATH, l984 ; SERVAIN e t a1.,1985; NICHOLSON et ENTEKHABI,1987). This r e l a t i o n could account f o r t h e coherence of t h e i n t e r a n n u a l SST v a r i a t i o n s and p a r t i c u l a r l y t h a t of t h e upwell ing throughout t h e Guinea Gulf, with p a r t i c u l a r yea r s or per iods having warmer ( co lde r ) s e a s u r f a c e waters MERLE,^^^^) (F ig .7 ) . Such synchronei ty on in t e rannua l s c a l e amplif ies

IO"

10"

O "

O

O.' ; .

O'

Figure 6 - Temperature d i s t r i b u t i o n of s u r f a c e water , s u r f a c e water c i r c u l a t i o n and upwell ing i n t h e t r o p i c a l A t l a n t i c Ocean f o r u. The currents (from nor th t o s o u t h ) : CC, Canary c u r r e n t ; DNA, North A t l a n t i c d r i f t : CNE, North e q u a t o r i a l cu r ren t ; CCEN, North e q u a t o r i a l coun te r cu r ren t ; CG, Current of Guinea; CSE, South e q u a t o r i a l c u r r e n t ( e q u a t o r i a l upwell ing) ; CCES, South e q u a t o r i a l c o u n t e r c u r r e n t ; DSA, South A t l a n t i c d r i f t ; CBI Benguela c u r r e n t (from WAUTHY,1983, fig. .20A). The

a r e shaded.

600 60 I

the impact of SST variations on the climate of the adjacent continent. An example of this relation between SST and climate can be

found during southern winter in the appearance of the "little dry season" (generally 2 months) in the rain forest north of the Guinea Gulf (DROCHON, 1976; BAKUN, 1978; HISARD, 1980) (Fig. 8,9) . This dry season, which is also the coldest period of the year (Fig.9), depends directly on the presence in. the lower troposphere of non-precipitating stratiform clouds, generated by the upwelling of cold waters at this time of the year (DROCHON, 1976). On the adjacent continent, the subsidence maintained aloft by the anticyclonic conditions spreading northward during southern winter explains the persistance of the stratiform clouds above the rain forest region (Fig.8). East of the Guinea Gulf, closer to the equator, the impact of the upwelling is larger because the same chain of oceanic and atmospheric anomalies prevails during four months. This four month dry season in the regions of Gabon and Congo is characterized by a quasi-permanent and non-precipitating stratiform cloud cover which extends at least 800 to 1000 km inland and reduces temperature and evaporation and, as a consequence, maintains a moist atmosphere (SAINT-VILI1977). Without this atmospheric humidity a four month dry season would have resulted in a replacement of the forest by savanna.

The relation between this dry season of southern origin and the lowering of SST is very clear and i s the most obvious phenomenon in the chain of atmospheric and oceanic anomalies. This conclusion is also supported by the observation that in some particular recent years (for example 1968 or 1984) warm waters persisted (Fig.7) because the upwelling were weak or absent. This phenomenon is similar to the "El Nino" of the eastern Pacific Ocean (HISARD, 1980; MERLE, 1980) . During the "warm" years this season was not dry whatever, but instead it was very wet with heavy rainfalls (HISARD,1980; GUILLOT,1985).

4.4.Major role of upwelling in the climatic changes of the rain forest region

The various present-day examples mentioned above (5 4.1 and 4.2) of the climatic action, either drying or (and) cooling of stratiform cloud covers, show that the generation of upwelling have a basic role in the climate system (cf. FLOHN,1982, 1983). Some authors have already shown that the upwelling of cold waters have influenced the palaeoclimates of Africa (VAN ZINDEREN BAKKER, 1982) or other regions (FLOHN, 1986) .

Figure 7 - Incerannual sea surface temperatures (SST) anomalies for some successive years of the recent past in the Guinea Gulf (Figure adapted from MERLE, 1980).

- Oblique shading : Positive anomalies ("warm waters") - Large dots : Negative anomalies from O to 0.5'C (upwelling,

- Small dots : Large negative anomalies of more than 0.5OC The recorded anomalies concern, from north to south : - m, a coastal locality, south of Ghana,and u (homogeneized NANSEN data, NOAA, Asheville,USA),

- latidunal strip of 10' longitude (O-lOoW) between 4' and 5'N along the northern shore of the Guinea Gulf,

- north of Mardsen square 300 (MS 300) ,an equatorial band 2' wide (equator - 2"s) between Oo and looW. This band corresponds to the equatorial upwelling area,

"cold waters").

(strong upwelling, "cold waters") .

- the entire Marsden square 300 (MS 300) (O-10'W; O-loos), - the entire Mardsen square 371 (MS 371) (O-lOOE; 1OoS-2O0S). , MFRLE(1980) note that the choice of these areas has been mainly determined by the density and the quality of the data. The coherence of anomalies throughout the Guinea Gulf is the principal feature of this figure. 4.4.1. The last phase of maximum aridity, about 18,000 yr BP

In the tropical Atlantic Ocean and particularly in the Guinea Gulf, the reconstruction of SST by the assemblages of plantonic foraminifera and radiolaria shows that at about 18,000 yr BP the equatorial cold waters upwelling maintained a nearly constant position, and were 4' - 8'C lower than those of the present southern winter (Fig.10) (PRELL et a1.,1976; MORLEY et HAYS,1979; MIX et a1.,1986). Further, PRELL et a1.(1976) have shown that during the southern summer (February), which

I 602

p----(-:-.-ig f ., . , ;. i . I , . ., ,.., '. . . 2 '.. ':? . .1 . . . . ( 1 . ~

F i g u r e 8 - Geographical distribution of southern winter dry season in equatorial Africa (adapted from LEROUX,1983). The length of this dry season increases from about 2 months in the north (=

"the little dry season") to more' than 4 months a year southwards. Note in West Cameroon the absence of this dry season, replaced by a pluvial paroxysm.

IAslDJAN CLIMATE DIAGRAM1

n --------+UPUELLIIIG +--------- r3

January February Haich A p r l l May June July August S e p t d e r October Hovanber December

Figure 9 - Diagram of seasonal climatic change in Abidjan (Ivory Coast), showing the succession of the main types of clouds (adapted from DROCHON 1976) and the principal elements Of climate : rain, air temperature, evapotranspiration and sea surface temperature (data from ORSTOM and ASECNA) .The Adijdan station is representative of the western sector of the rain forest area.

F i g u r e 10 - Sea surface temperatures estimates (August and February) according to tranfer functions of foramineferal assemblages in the north tropical Atlantic Ocean during the last glacial maximum, ca. 18,000 yr BP. Present-day values are provided for còmparison*'*'(from 'PRELL et al., 1976, fig.11 and 12, in Geol,

145).

605

is the season of warmest water, the surface temperatures were 3 degrees- lower than today.This suggests that the upwelling of cold waters would have been a nearly year-round feature. From what we know from the present-day atmospheric phenomena related to upwelling, these data mean also a very powerfull anticyclonic high pressure above the Southern Atlantic associated with very strong trade winds. Indeed, research in this field has shown an important strengthening of trade winds around 18,000 yr BP (NEWELL et a1.,1981). In conclusion the present-day interannual synchronism throughout the upwelling areas in the Guinea Gulf (equatorial and coastal) (Fig.7) means that in the past the SST variations have had a large climatic impact on the equatorial regions of Africa, mainly those covered today with rain forest. These phenomena and the action of stratiform cloud cover inland could explain (1) the aridification of the climate and the disappearance of large areas of rain forest and (2) the cooling effect of these quasi-permanent cloud covers and the accompanying extension of montane biotopes to lower elevations.

4.4.2.The abrupt change near 9000 years BP and the reappearance of the lowland rain forest

In West Africa, pollen evidence and other' data from the Bosumtwi lake (5 2.1) show that the near absence of forest and the extended spreading of montane vegetation to low elevations lasted until about 9000 yr BP. This coincided .with important warming at higher latitudes in both hemispheres which began about 15,000 yr BP (LORIUS et a1.,1979; BARD et a1.,1987; SAFWTHEIM et al., 1981). Pollen data (ROCHE, 1979; CARATINI et GIRESSE, 1979; M'BENZA-MUAKA et ROCHE, 1980) and geological data (DE PLOEY, 1969; GIRESSE et LANFRANCHI, 1984; PEYROT et LANFRANCHI, 1984; KADOMURA et a1.,1986) from the Zaïre river catch- ment, the western Congo and southern Cameroon, show a comparable history of lowland rain forest reappearance. The foraminiferal data of PRELL et al. (1976) suggest that in the belt of equatorial upwelling, SST values at 9000 or 10,000 yr BP became close to those of the present. This change coincided with the rain forest reappearance and climate warming (disappearance of the montane element).

The increase in temperature from ca. 15,000 to 9000 yr BP affected mainly middle and high latitudes but in tropical regions and particularly in the African rain forest, the change was small. As explained above (5 4.3), this phenomenon is probably related to the persistence of cold waters upwelling throughout the Guinea Gulf. The SST outside the upwelling areas increased (MIX et a1.,1986), which provided more water vapor for the monsoon and resulted in more cloud covers inland and probably an increase in rainfall . At the time of the rain forest reappearance, the level of Bosumtwi lake was already very high above the present-day level (Fig.2,B). Since lower temperatures prevailed at that time (presence of montane vegetation until ca. 9000 yr BP), we could infer that the high lake level was partly caused by low evaporation rates, but also by increase in precipitation. These deductions indicate that the reappearance of the rain dependent on rainfall and forest near 9000 yr BP was not solely

606

. .

607 ..*

humidity, which were ve ry s u f f i c i e n t s e v e r a l m i l l e n i a b e f o r e t h i s da t e , b u t a l s o on temperature and c h i e f l y amount of sunshine. For t h i s reason, w e b e l i e v e t h a t t h e reappearance of t h e r a i n f o r e s t was mainly governed by t h e v a r i a t i o n s of c loud covers and cloud types.

One cou ld e s t i m a t e t h a t between 15,000 and 9 0 0 0 y r BP t h e present-day July- type weather with major r a i n produced by massive n imbos t r a tus c louds (Fig.91, had p r o g r e s s i v e l y dominated each yea r du r ing s e v e r a l months b e f o r e J u l y . Becausk w e know t h a t t h e lowland r a i n f o r e s t s r e q u i r e l a r g e s u n l i g h t i n t e n s i t y d u r i n g many months (LEIGH, 1975: WALTER e t BRECKLE, 1985; e tc) , t h e reappearance of t h e r a i n f o r e s t cou ld be due t o a s u b t l e balance between t h i s July- type weather and t h e March t o May-type weather which is c h a r a c t e r i z e d by longer pe r iods of sunshine ( F i g . 9 ) .

4.4.3. The Holocene pe r iod between about 9000 and 4000 yea r s BP

The SST d a t a f o r t h e Holocene o f PRELL e t al. (1976) a r e n o t s u f f i c i e n t l y p r e c i s e t o detect and date ' temperature v a r i a t i o n s , bu t some o t h e r i n d i r e c t d a t a show t h a t SST anomalies have had an important impact on t h e c l ima te and t h e r a i n fores t .

Near t h e c o a s t of wes te rn Congo widespread savannas a r e p r e s e n t l y found: however DECHAMPS e t a l . (1988). and SCHWARTZ e t a l . ( t o be publ ished) have s t u d i e d s e v e r a l e x t e n s i v e p a l a e o s o l s with i n s i t u remains of numerous tree stumps which belong t o a r a i n f o r e s t f l o r a . Seve ra l t ree stumps and a l s o some samples of o rgan ic m a t t e r from t h e p a l a e o s o l s w e r e radiocarbon da ted . A first series of 1 0 samples w e r e d a t e d from about 6500 t o 3100 y r BP and!a second series of 3 samples were d a t e d about 500 t o 600 y r BP. Because t h e present-day savannas of t h i s r eg ion a r e mainly r e l a t e d t o a l ong dry season (MAKANY,1964) which a r e caused by t h e c o l d water upwelling (SAINT-VILI1977) (§ 4 . 3 ) , t h e i r replacement by r a i n f o r e s t was probably caused by a suppression o r reduct ion of upwelling i n t h i s a r ea .

Other d a t a r e l a t e d t o t h e upwelling suppression o r r educ t ion i n t h e Gulf o f Guinea d u r i n g p a r t s of Holocene t i m e can b e , ob ta ined by t h e s t u d y of t h e Bosumtwi l a k e sed imen t s . Today, t h i s l a k e i s s t r a t i f i e d with anoxic wa te r s below about 10m, with a maximum water depth of about 80m (BEADLE, 1974) , bu t almost each yea r an ove r tu rn of deep anox ic wa te r s o c c u r s . Because LIVINGSTONE ( p r i v a t e commun.) counted about one l amina t ion f o r one radiocarbon y e a r i n t h e co re c o l l e c t e d i n t h e l a k e ( F i g . 2 ) , t h e l a m i n a t i o n s of t h e sediment correspond t o t h i s y e a r l y phenomenon. The ove r tu rn i s mainly r e l a t e d t o a lowering of a i r temperature i n August and September which i s induced by t h e non-precipidat ing s t r a t i f o r m cloud covers gene ra t ed by t h e upwell ing (5 4.3) ( F i g . 9 ) . This c o o l i n g a l s o a f f e c t s t h e water column which a t t h i s t i m e becomes more o r less homothermal (BEADLE,1974). Because, a t t h e same t i m e , t h e wind stress from t h e sou thwes te r l i e s i s a t i ts maximum, t h e i n s t a b l e waters a r e overturned (BEADLE,1974; WHYTE,1975). But f o r t h e co re (Fig.2,A), a l a r g e p a r t of t h e Holocene sediment i s unlaminated and of Sapropel t y p e between about 9 0 0 0 y r BPI when t h e r a i n f o r e s t reappeared, and about 3700 y r BPI when t h e laminat ions reappear suddenly (TALBOT et a1. ,1984) . So,

t h i s unlaminated i n t e r v a l of t h e c o r e and t h e high l a k e l e v e l of t h i s p e r i o d ( a t maximum 105m above present-day l a k e level) can be r e l a t e d t o an absence o r r educ t ion of t h e " l i t t l e d r y season". E i t h e r an annual p r o l o n g a t i o n through August and September of t h e June-type weather (F ig . 9 ) which i s c h a r a c t e r i z e d by heavy p r e c i p i t a t i o n from n i m b o s t r a t u s c l o u d s , o r t h e a n n u a l e a r l y s t a r t of t h e October-November-type wea the r ( F i g . 9 ) which numerous t o w e r i n g . cumuliform clouds a r e t y p i c a l , probably occurred. For sedimentological reasons ( i n West Af r i ca f l u v i a t i l e sediments show a t r a n s i t i o n from c l a y t y p e d e p o s i t s t o m o r e s a n d y t y p e n e a r 7 0 0 0 y r BPI cf.MALEYI1981,p.519 : MALEY,1982,1983) w e e s t i m a t e t h a t t h e June-type w e a t h e r domina ted u n t i l a b o u t 7 0 0 0 y r BP a n d a f t e r , t h e October-November-type weather (Fig.9) dominated u n t i l about 3700 y r BP. The y e a r l y r a i n f a l l was c e r t a i n l y h i g h e r t h a n today ( ca . 1520 mm) ( 5 4 . 4 . 4 1 , but , because t h e po l l en d a t a f o r t h i s p e r i o d e x h i b i t no p a r t i c u l a r change, t h e composition of t h e f o r e s t must have remained o f - a semi-deciduous t y p e (MALEY et LIVINGSTONE,1983, and unpubl ished d a t a : s p e c t r a have on ly very few p o l l e n g r a i n s of Caesa lp in i aceae - c h a r a c t e r i s t i c of t h e e v e r g r e e n f o r e s t t y p e , see § 2.2 and BRENAC, 1988) . Consequently one concludes t h a t d u r i n g t h i s p e r i o d t h e no r the rn w i n t e r d r y season had t h e same l e n g t h as today, i .e. t h r e e months (December t o February) ( F i g . 9 ) . P r e s e n t c o n d i t i o n s i n W e s t Cameroon may be r e p r e s e n t a t i v e of t h o s e f o r t h e p e r i o d 9000 y r BP t o about 3700 y r BP i n t h e Bosumtwi r eg ion and probably a l so f o r o t h e r p a r t s of t h e Afr ican r a i n f o r e s t . I n West Cameroon indeed, t h e t h r e e months from J u l y t o September r e p r e s e n t t h e r a i n i e s t s eason of t h e y e a r (F ig .8 ) ,and a r e d e s c r i b e d a s a " p l u v i a l paroxysm" by SUCHEL (1972).

4 . 4 . 4 . The r eg res s ion of t h e l a k e Bosumtwi du r ing l a t e Holocene-time . -

A f t e r 9000 y r BPI t h e present-day i n t e r r u p t i o n of t h e r a i n f o r e s t a r ea which i s c a l l e d t h e "Dahomey Gap" was p robab ly a b s e n t , because of h ighe r humidity ($5 4.4.3), and f o r b iogeograph ica l r easons (see t h e d i s c u s s i o n i n MALEY, 1987) . The Dahomey Gap p robab ly appeared between 4 0 0 0 and 3000 yea r s ago, when l a k e Bosumtwi a b r u p t l y regressed.

The abrupt r eg res s ion of t h e l ake , more than 120m, is da ted a t about 3700 y r BP which i s t h e d a t e of t h e sudden reappearance of t h e l amina t ions (TALBOT e t a1. ,1984), and i s t o be a s s o c i a t e d with t h e reappearance of t h e " l i t t l e d ry season" i n l a n d and t o t h e upwelling of c o l d water i n t h e Guinea Gu l f . However t h e p o l l e n d a t a fra" t h i s . p e r i o d show t h e p e r s i s t a n c e of t h e r a i n f o r e s t (Fig.2 and 3 ) . I n . . orde r t o exp la in t h i s , t h e present-day water budget OZ t h e l a k e which i s s l i g h t l y p o s i t i v e (average r a i n f a l l on t h e c r a t e r catchment i s 1 5 2 0 m / y e a r w i t h e v a p o r a t i o n a t o n l y 1450mm/year - TALBOT e t DELIBRIAS, 1977) should be considered. There i s an average excess o f 70" which e x p l a i n s t h e present-day t r a n s g r e s s i v e phase of t h e l a k e . which has a p p a r e n t l y been underway f o r s e v e r a l c e n t u r i e s (TALBOT e t a1.,1984, f i g . 6 ) , because a l a c u s t r i n e minimum probab ly occur red i n 15 th cen tu ry AD (MALEY, i n shed.) . This r e g r e s s i o n phase, l i k e t h e one da ted between 4000 and 3 0 0 0 yr BP, was caused by a t i e t e r i o r a t i o n of

' 608

the lacustrine water budget that linked a rainfall decrease to a probable evaporation increase. If this rainfall decrease was distributed over the same number of months, then, provided the total precipitation does not reach values below 1200mm/ year, rain forest could have remained, based on observations that today some northern parts of rain forest in Ghana and Ivory Coast receive yearly rainfall in this range (Fig.1). Thus, provided minima are not below 1200mm/year, the essential factor to maintain the rain forest in late Holocene time is good distribution of rainfall throughout the year, rather than the annual total.

5 . CONCLUSIONS

After having briefly introduced the problem of rain forest refugia during the last glacial maximum and shown the pollen data which clearly demonstrate that the African rain forest area has been fragmented indeed and survived the glacial period in some refugia, one had attempted to reconstruct the weather types which may account for the,recorded vegetational and lacustrine variations. Particular attention has been paid to the climatic reconstruction (1) during the phase of fragmentation of the rain forest belt and the extension of montane biotopes to lower elevation, (2) during the forest recolonization and its extension in early and middle Holocene time, and ( 3 ) the return to its present-day limits during the last four millenia and particularly the opening of the "Dahomey Gap". The latter phenomenon is probably related to a deterioration of the water budget, which is also responsible for the regressions of Lake Bosumtwi.

The question of why forest refugia have been formed and how were climatic conditions there, remains without precise answer. One of the major proposed refugia is located in West Cameroon (Fig.1) (MALEY, 1987) , where today climatic conditions are atypical because this area experiences no "little dry season" during southern winter (Fig.8). For this reason the yearly rainfall is very high. This particularity could account for the survival of the tropical rain forest during dry periods. However, we see that there have been other refugia in regions which experience today a dry season during southern winter such as in Gabon and southern Cameroon. This suggests that other factors are necessary to explain the geographical position of

The study of climatic conditions of West Cameroon could provide the necessary informations to allow the development of a hypothese. The exact reasons of the present-day July to September pluvial paroxysm of West Cameroon are not well understood (SUCHEL,1972) and could be-the resu_lt.\of several causes. It could be related first to the geographic position at the far end o f the Guinea Gulf where today the SST seem to be unaffected by the upwelling phenomenon or to the presence of high mountains with possible orographic effects and strong dynamic lifting associated with upper air circulation.

, - As ,we have seen above (S 2 and 31, during colder periods, lowland rain forest and montane vegetation were probably close

rEfPgip: - * i 4 -

609

together such as in West Cameroon. Presently, as hypothesis, one suggests the transformation of non-precepitating stratiform clouds to precipitating nimbo-stratus or cumuliform clouds through orographic or dynamic lifting of air masses (MALEY,1982; see examples of physical processes in LUDLAM, 1966; etc) . In this way, locally, prevailing higher temperatures and rainfall necessary for lowland rain forest to survive could have been achieved.

More detailed climatological studies in West Cameroon and in restricted areas where today montane biotopes extend to low elevations, may bring new data leading to a better understanding of the African rain forest palaeoclimatology.

ACKNOWLEDGEMENTS

This paper is a contribution to a research program on the Palaeoenvironments of the African rain forest and particularly of West Cameroon. The program results from collaboration between several Cameroonian scientific Institutions (MESRES, ISH, CGN, CRH, IRGM, Yaounde University), ORSTOM, France (J.MALEY, Research Unit A3 and GEOCIT Program) and the Zoology Department of Duke University, USA (D .A.LIVINGSTONE) . For helpful discussions, thanks are due to J.L. AMIET (Zoology Dept. , Yaounde Univ. ) , R. LETOUZEY (Museum Hist. Nat., Paris) , D. A. LIVINGSTONE (Duke Univ. , Durham, USA) , S .MORIN (Geography Dept.,Yaound& Univ.) and D.W.THOMAS (Missouri Botanical Garden and IRA, Kumba S W P ) . The author is also grateful to W.L.PRELL (Lamont Geological Observ.,New-York) for permission to reproduce material published in -1 Societv of &neahaJt5alUUL * (1976, ~01.145, fig.11 and 12). Thanks are done also to two anonymous reviewers for constructive remarks and to J. SCHUHMACHER (Mineralogisches Inst. , Kiel Univ. ) who greatly improve the English. The cores from Lake Bosumtwi, Ghana and Lake Barombi Mbo, Cameroon, were collected by crews led by D.A.LIVINGSTONE. The funding of coring operations in Cameroon was. provided by ORSTOM, CNRS, NSF and Duke University, and in Ghana by NSF.

KEYWORDS

Africa - Lake Bosumtwi, Ghana - Lake Barombi Mbo, West Cameroon. - Lowland Rain Forest - Montane vegetation - Pollen Analyses - Forest Refugia - Biogeography - Speciatian - Dahomey Gap - Upper Quaternary - Climatic Change - Palaeoclhatology - Tropical Atlantic - Sea Surface Temperatures - Upwelling - Cloud - Stratiform clouds -

610 61 1

REFERENCES

ACHOUDONG, G. 1985, Etude dcologique et floristique de la vegetation des collines de Yaounde au-dessus de 1000m. These Univ.Yaound6, Ecologie V6g&., 301 pp.

AIRY-SHAW, H.K. 1947, The vegetation of Angola. J.of135,23-48. AUBREVILLE,A. 1949, Contribution d la paleohistoire des forgts de

AUBREVILLE,A.1962, Savanisation tropicale et glaciations quaternaires.

BAKUN, A. 1978, Guinea current upwelling. Nature, 271,147-150. BARDIE., ARNOLD,M., DUPRAT,J., MOYES,J. et DUPLESSY,J.C. 1987,

Reconstruction of the last deglaciation : deconvolved records of 6 180 profiles, micropaleontological variations and

101-112. BAYNTON,H.W. 1968, The ecology of an elfin forest in Puerto Rico. 2.

The microclimate of Pico del Oeste. J.of Arnold, 49,

BEADLE, L.C. 1974, The inland waters of Africa. An introduction to

BERNARDI, G. 1979, Contribution d la biogeographie des montagnes

BOOTH, A.H. 1958, The Niger, the Volta and the Dahomey Gap as geographic barriers. Evolution, 12,48-62.

BRENAC, P. 1988, Evolution de la vegetation et du climat dans l'Ouest Cameroun entre 25.000 et 11.000 ans BP. Actes Xeme Symposium Ass.Palynologues Langue Française,Trav.Sect.Sci.& Tech.=

CARATINI,C., et GIRESSE,P. 1979, Contribution palynologique d la connaissance des environnements continentaux et marins du Congo ?i la fin du Quaternaire. C.R.Acad.Sc.,sBrie D,288,379-382.

CARCASSON,R.H. 1964, A preliminary survey of the zoogeography of

CLAYTON, W.D. 1976, The chorology of African mountain grasses. X.a f ayIl., 31, 273-288.

COETZEE,J.A. 1964, Evidence for a considerable depression of the vegetation belts during the Upper Pleistocene on the East African mountains. Nature,204,564-566.

DECHAMPSIR. , GUILLET,B. et SCHWARTZ,D. 1988, Ddcouverte d'une flore forestiere mi-Holoc8ne (5800-3100 BP) conservee in situ sur le littoral pontenegrin (R.P. du Congo) .C.R.Acad.Sc. , serie II, 306,

DE PLOEY,J. 1969, Report on the Quaternary of western Congo.

DROCHONIA. 1976, Donndes climatologiques au sol et en altitude Rour la station d'Abidjan. Note ASE- (Agence Securite Naviga. Abrienne) ,55, 14 pp.

ELENGA,H. 1987, Les Plateaux Bateke (Congo). Paldoenvironnements Quaternaires d'apres l'dtude du sondage du bois de Bilanko.

l'Afrique tropicale. Soc.Edit.Bboar.Marit. et C o a I 99 PP-

Adansonia, 2,16-84.

accelerator mass spectometric 14C dating. Climate D- I 1,

419-430.

tropical limnology. Longman, London, 365 pp.

africaines. I. Gbn6ralitBs. C&.Soc.BioaBoar. 479,6-10.

s P-, 25,91-103

African butterflies. . ,2,122-151.(

615-618.

of Africa, 4,65-68.

1 Ftudes AoDrofondies, Universite d' Aix-Marseille II, 41 PP.

ENDLER,J.A. 1982, Pleistocene forest refuges: fact or fancy. in G.T.Prance (ed) Bioloakal diverslflcation in the tropic& Columbia Univ.Press, New-York, pp.641-657.

Ranunculaceae - Malvaceae. mt.Bota.CoimBra . 176 pp. FEDOROV,A.A. 1966, The structure of the tropical rain forest and

speciation in the humid tropics. J.Ecoloay, 54,l-11. FLENLEY, J.R. 1979, The equatorial rain forest : a geological history.

Butterworths, London, 162 pp. FLENLEY, J.R. 1984, Late Quaternary changes of vegetation and climate

in the Malesian Mountains. Erdwissensch. Forsch -, Wiesbaden, 18,

FLOHN,H. 1982, Oceanic upwelling as a key for abrupt climatic change.

FLOHN,H. 1986, Singular events and catastrophes now and in climatic

GIRESSE,P. et LANFRANCH1,R. 1984, Les climats et les oceans de la Region Congolaise pendant 1'Holocene. Bilans selon les dchelles et les methodes de l'observation. &.l.aeoecOav of Afrlca ' ,161

GRUBB, P. J. 1971, Interpretation of the "Massenerhebung" effect' on tropical mountains. -,229,44-45.

GRUBB,P.J. 1974, Factors ccontrolling the distribution of forest-types on tropical mountains: new facts and a new perspective. in J.R.Flenley (ed),-on in Malesla , Trans.3rd Aberdeen-Hull Symp. on Malesian Ecology, pp.13-45.

GRUBB,P.J. 1977, Control of forest growth and distribution on wet tropical mountains , with special reference to mineral nutrition. aM.Rev.Ecol.Syst.,8,83-107.

GRUBB,P.J. et WHITMORE,T.C. 1966, A comparison of montane and lowland rain forest in Ecuador. II.The climate and its effects on the distribution and physiognomy of the forests. J.Ecolocrv, 54, 303-333.

GUILLOT,B. 1985, Temperature de surface de la mer et pluviosite autour * * , Lannion, 6, du Golfe de Guide. 5

HAFFER,J. 1982, General aspects of the refuge theory. in G.T.Prance , Columbia (ed), FLaLuical diverslfrcation in the t r d

Univ.Press, New-York, pp. 6-24. HAMILTON,A.C. 1973, The history of the vegetation. in E.M. Lind et

M.E. S. Morrison (eds) , T h e o n of ' , Longman, London, pp.188-209.

HAMILTON,A.C. 1982, Environmental history of East Africa. A study of the Quaternary. Academic Press, London, 311 pp.

HALL, J.B. 1973, Vegetational zones on the southefn slopes of -,Mount Cameroon. Vwe- ' ,27,49-69.

HALL, J.B. et SWAINE,M.D. 1981, Distribution and Ecology of vascular plants in a tropical rain forest : Forest vegetation in Ghana.

. .

EXELL,A.W et MENDONCA,F.A.1937, Conspectus Florae Angolensis. I,

261-267.

W t e o . of Jagan,60,268-273.

history. Naturwissenschaften , 73,136-149.

77-88.

* .

13-14.

. .

' 612

1

613

W.Junk, The Hague, 383 pp. HASTENRATHIS. 1984, Interannual variability and annual cycle :

mechanisms of circulation and climate in the tropical Atlantic

HEPPER,N. 1963, Flora of West tropical Africa. in J.Hutchinson et J.M.

Dalziel (ed),Crown Agents for Oversea G o v e . , London, vo1.2.

HIRST,A.C. et HASTENRATH, S. 1983 (a) , Diagnostics of hydrometeorolo- gical anomalies in the Zaïre (Congo) basin. 0uat.J.R.Met. Soc., 109,881-892.

HIRST,A.C. et HASTENRATH, S. 1983 (b) , Atmosphere-Ocean mechanisms of climate anomalies in the Angola-Tropical Atlantic sector.

HISARD,P. 1980, Observationn de reponse de type 'El Nino' dans l'Atlantique tropical oriental, Golfe de Guinee. Oceanoloaica &-ta, 3,69-78.

HOFLICH,0.1972, Die meteorologischen Wirkungen kalter Auftriebswasser- gebiete. Geofor~,11,35-46.

I 2, HOWARD,R.A. 1970, The "Alpine" plants of the Antilles.

KADOMURA,H., HORIIN., KUETE,M., TAMURAIT., OMI,G., HARUK1,M. et CHUJO, H. 1986, Late Quaternary environmental changes in southern Cameroon: a synthesis. in H.Kadomura (ed), Environmental m u e s in t r o w . w e studies in -. Hokkaido Univ.,pp.145-158.

LEIGH,E.G. 1975, Structure and climate in tropical rain forest. Annual

LER0UX.M. 1983, Le climat de l'Afrique tropicale. Champion, Paris, 633

LETOUZEY, R. 1968 , Etude phytogeographique du Cameroun. mcvclo-

LETOUZEY,R. 1978, Notes phytogeographiques sur les Palmiers du

LETOUZEY, R. 1985, Notice de la carte phytogeographique du Cameroun au 1/500.000 -.Carte -eta a, Toulouse & m a A Bsr;M. , Yaounde.

LIVINGSTONE,D.A.et CLAYTON,W.D.1980, An altitudinal cline in tropical African grass floras and its paleoecological significance. Q!da~&~.l3, 392-402.

LORIUS,C., MERLIVAT,L., JOUZEL, J. et PGURCHET,M. 1979, A 30,000 yr isotope climatic record from Antarctic ice. Nature, 280,

sector. m l v We- Rev h, 112,1097-1107.

J . P h v s , O c ~ . , 13, 1146-1157.

24-28.

& Svstematics, 6,67-86.

pp. & Atlas, 250 maps.

-, Paris, 49, 508 pp.

Cameroun. Adansonla ,18,293-325.

644-648. LUDLAM,F.H. 1966, Cumulus and cumulonimbus convection. u, 18,

687-698. MAKANY,L. 1964, La côte Atlantique du Congo .: cadres géographiques et

geologiques, leur influence sur la repartition de la vegetation et sur les possibilites agricoles du territoire. Actes

MALEY,J. 1981, Etudes palynologiques dans le bassin du Tchad et paleoclimatologie de l'Afrique nord-tropicale de 30.000 ans d l'&poque actuelle. Trav.& Doc- ORSTm , Paris,l29, 586 pp.

pp. 891-907.

MALEY

MALEY

MALEY,J. 1982, Dust, clouds, rain types and climatic variations in tropical North Africa. Duat .-. ,

MALEY,J. Histoire de la vegetation et du climat de l'Afrique nord-tropicale au Quaternaire rdcent. Bothalla 14 (3-4) I 377-389.

MALEY,J. 1984, Influence des nuages sur les paleoenvironnements : quelques exemples SC.-, pris Bordeaux, dans le p.373 Quaternaire (Abstract). africain.

MALEY,J. 1987, Fragmentation de la for& dense humide africaine et extension des biotopes montagnards au Quaternaire recent : nouvelles donnees polliniques et chronologiques. Implications paleoclimatiques et biogeographiques. -UV of AfriCa, 18,307-334.

J. 1988, L'importance de la tradition orale et des donnees historiques pour la reconstitution paldoclimatique du dernier millenaire sur l'Afrique nord-tropicale, (manuscrit),6 pp.

J. et BRENAC,P. 1987, Analyses polliniques prelimhaires du Quaternaire recent de l'ouest Cameroun : mise en dvidence de refuges forestiers et discussion des problbmes paleoclimatiques

MALEY, J. , CABALLE,G. et SITA,P. 1988 (a), Etude d'un peuplement residue1 B basse altitude de Podocarpus latifolius sur le flanc Congolais du Massif du Chaillu. Implications palCoclimatiques et biogeographiques. Etude de la pluie pollinique actuelle. In D.Schwartz et R.Lanfranchi (ed) , pavsaaes 0-s de

Tray.& Do-, 14 pp. (to be

18,l-16. 1983, *

' , 17 , pp. 129-142. j

published). MALEY. J., GIRESSE,P., THOWENY, N. , BRENAC, P. I KELTS,K. , LIVINGSTONE I . .

D . A . , KLING, G., STAGER, C. , HAAG, M. I FOURNIER, M. I BANDET,Y. WILLIAMSON, D. et ZOGNING,A. 1988 (b) , Lithostratigraphy, Volcanism, Paleomagnetism and Palynology of Quaternary lacustrine deposits from Barombi Mbo (West Cameroon) : preliminary results. J.Voloano. and Geoth.Res -, 18 pp. (to be published).

MALEY,J. et LIVINGSTONE,D.A.1983, Extension d'un element montagnard dans le sud du Ghana (Afrique de l'Ouest) au Pleistoc&ne superieur et h l'Holocbne inferieur : premieres donnees polliniques. C.R.Acad.Sc.,sBxie 2,296, 1287-1292.

MAYRIE. et O'HARA,R.J. 1986, The biogeographic evidence supporting the Pleistocene forest refuge hypothesis. Evolution ' , 40,55-67.

M'BENZA-MUAKA et ROCHE,E. 1980, Exemple d'bvolution paleoclimatique au PlBistoc&ne terminal et h l'Holoc&ne au Shaba (Zaïre). Symposium Rot-, "Palynologie 27,137-148. et Climats", MkmAU3eum Nat. k & & d d L ~

MERLE,J. 1980, Variabilite thermique annuelle et interannuelle de l'Oc8an Atlantique equatorial est. L'hypothgse d'un "El Nino" Atlantique. Acta ,3,209-220.

MIX,A.C., RUDDIMAN,W.F.~~ M C I N T Y R E ~ A . ~ ~ ~ ~ , Late Quaternary paleoceano- graphy of the tropical Atlantic. 2: The seasonal cycle of sea surface temperatures, O-20,000 years B.P. 1,339-353.

MOREAU, R.E. 1935, A synecological study of Usambara, Tanganyika

614 615

Territory, with particular reference to birds. J.Ecolouy, 23, 1-43.

MOREAU,R.E. 1938, Climatic classification from the standpoint of East African biology. ,L&~Lgy,26,467-496.

MOREAU,R.E. 1963, The distribution of tropical African birds as an indicator of past climatic changes. in F.C.Howel1 e t F.Bourli&re (ed) ,African eco- * , Aldine Press, Chicago, pp.28-42.

MOREAU,R.E. 1966, The bird faunas of Africa and its islands. Academic Press, New-York, 424 pp.

MORLEY,J.J. et HAYS,J.D. 1979, Comparison of glacial and interglacial oceanographic conditions in the south Atlantic from variations in calcium carbonate and Radiolarian distributions. Quat.&., 12,396-408.

MORLEY,R.J. et FLENLEY,J.R. 1987, Late Cainozoic vegetational and environmental changes in the Malay archipelago. in T.C.Whitmore (ed), &ioaeoaraDhical Evolution of U a v Ar-. Clarendon Press, Oxford, pp.50-59.

MORTON,J.K. 1968, Sierra Leone. in I.& 0.Hedberg (ed), Conservati- Qf veaetation in Africa south of the S a m . A c t a Phvtoaeo-

NEWELL,R.E., GOULD-STEWART,S.et CHUNG,J.C. 1981, A possible interpre- tation of paleoclimatic reconstructions for 18,000 BP for the region 60°N to 6OoS, 6OoW to 100°E. Ueoecolouv of A f r h , 13,l-19.

NICHOLSON,S.E. et ENTEKHAB1,D. 1987, Rainfall variability in equatorial and southern Africa: relationships with sea surface temperatures along the southwestern coast of Africa. J.Climate &., 26,561-578.

PALMER,P.G. 1982, A stratigraphic analysis of fossil grass cuticles from lake Bosumtwi, Ghana, West Africa and its paleoecological significance. XInd IntW.Ouat..Moscoy , Abstracts, I, p.244.

PEYROT,B. et LANFRANCH1,R. 1984, Les oscillations morphoclimatiques recentes dans la vallee du Niari. -oav of Afrim, 16,265-281.

Columbia Univ. Press, New-York, 714 pp. PRELL,W.L., GARDNER,J.V., BE,A.W.H. et HAYS,J.D. 1976, Equatorial

Atlantic and Caribbean foraminiferal assemblages, temperatures and circulation : interglacial and glacial comparisons. G e o l o w S o c . A m e r i c a , l 4 5 , 247-266.

RICHARDS,P.W. 1963, Ecological notes on West African vegetation. 1I.Lowland forest of the southern Bakundu Forest Reserve. J. Ecola, 51,123-149.

RICHARDS,P;W. 1969, Speciation in the tropical rain forest and the concept of the niche. Biol.J&nn..Soc . , 1,149-153.

ROCHE,E. 1979, Vegetation ancienne et actuelle de l'Afrique centrale. can Ec- I 7,30-37.

SAINT-VIL, J. 1977, Les climats du Gabon. -v.Nat .- , 1,101-125.

Ca SueC.54,72-74.

PRANCE,G.T. (ed) , 1982, -al diversmcation . . in the Trow&.

I .

SARNTHEIN,M. , ERLENKEUSER,H. , et ZAHN,R. 1982, Termination I: the response of continental climate in the Subtropics as recorded in deep-sea sediments. .Gbol.-in 393-407.

SCHNELL,R. 1950, La for& dense. Introduction A 1'8tude botanique de la region forestiere d'Afrique occidentale. P.Lechevalier (ed),Paris, 330 pp.

SCHNELL,R. 1977, Introduction à la phytogeographic des pays tropicaux. 4 : La flore et la vegetation de l'Afrique tropicale. Gauthier-Villars (ed), Paris, 378 pp.

SCHWARTZ,D., GUILLET,B. et DECHAMPS,R. 1988, Etude de deux flores foresti&res mi-Holoche (6000-3000 BP) et subactuelle (500 BPI conservees in situ sur le littoral Pontenegrin (Congo). in D.Schwartz et R.Lanfranchi (ed), pavsaues 0 u a t e r n a i . d e

, ~ r a v . et Doc!mf!.nts ORSTQM,16 pp. (to be published).

SERLE,W. 1964, The lower altitudinal limit of the montane forest birds of the Cameroon mountains, West Africa. Bull.Brltlsh O r n i t h o . ur 84,87-91.

SERVAIN, J., PICAUT, J. et BUSALACCH1,A.J. 1985, Interannual and seasonal variability of the tropical Atlantic Ocean depicted by sixteen years of sea-surface temperature and wind stress. in ~.c.~.Nihoul (ed), -e - m o d e l s . Elsevier, Amsterdam, pp.211-237.

SUCHEL,J.B. 1972, La repartition des pluies et les regimes pluviometriques au Cameroun. Trav.& Documents G&o- m, Bordeaux, 5, 287 pp.

SUGDEN,A.M. 1982, The vegetation of the Serrania de Macuira, Guajira, Colombia: a contrast of arid lowlands and an isolated cloud forest. J.Arnold., 63,l-30.

TALBOT,M.R. et DELIBRIAS,G. 1977, Holocene variations in the level of Lake Bosumtwi,Ghana. Nature,268,722-724.

TALBOT,M.R. et KELTS,K. 1986, Primary and diagenetic carbonates in the anoxic sediments of Lake Bosumtwi, Ghana. GeOloQV, 14,912-916.

TALBOT, M. R., LIVINGSTONE, D. A . , PALMER, P. G., MALEY, J. , MELACK, J.M., DELIBRIAS, G. et GULLIKSEN, S. 1984, Preliminary results from sediment cores from Lake Bosumtwi,Ghana. Enlheoecoloav of Africar16,173-192.

THOMAS,D.W. 1986, Vegetation in the montane forests of Cameroon. in S .N.Stuart (ed) ,Conservation of Cameroon M o n t a n e t Report of Int.Counci1 Bird Preserv.,Cambridge,UK, pp.20-27.

VAN ZINDEREN BAKKER,E.M. 1982, African palaeoenvironments 18,000 years B.P .palaeoecoloav of Af 15,77-99.

VAN ZINDEREN BAKKER,E.M. et CLARK, J.D. 1962, Pleistocene climates and cultures in north-eastern Angola. Nature,196,639-642.

VAN ZINDEREN BAKKER,E.M. et COETZEE, J.A. 1972,A re-appraisal of late Quaternary climatic evidence from tropical Africa.meoecolQ,gy Gf Af~,7,151-181.

WALTER,H. et BRECKLE,S.W. 1985,Ecological systems of the Geobiosphere. 1,Ecological principles in global perspective. Springer-Verlag, Berlin.

,31, . .

. .

a -

l

i

?

616

WAUTHY,B. 1983, Introduction B la climatologie du Golfe de Guide. e troDule. m, 19,103-138.

WEBSTER,P.J. et STRETEN,N.A. 1978, Late Quaternary Ice Age climates Of tropical Australasia: interpretations and reconstructions. Quat.Ra.10,279-309.

WHITEIF. 1978, The Afromontane region. in M. J.A.Werger (ed) hy and Ecolouv of sou-, W. Junk, The Hague,

2 , 4 63-513. WHITEIF. 1981, The history of the Afrbmontane archipelago and the

scientific need for its conservation. A€-=, 19,33-54. WHITE, F. 1983, The vegetation of Africa. UNESCO/AETFAT/UNSO, Maps L

Memoir, 356 pp. WHITEIF et WERGER,M. J. 1978, The Guineo-Congolian transition to

southern Africa in M.J.A.Werger (ed) ,- and Ecology Qf so-. W. Junk The Hague, 2,601-620.

WHITMORE,T.C. 1975, Tropical rain forests of the Far East. Clarendon Press, Oxford.

WHITMORE,T.C., FLENLEY, J.R. et HARRIS,D.R. 1982, The tropics as the norm in biogeography ? Geoar.J.148,8-21.

WHITMORE,T.C. et PRANCE,G.T. 1987, Biogeography and Quaternary History in tropical America. Clarendon Press, Oxford, 214 pp.

WHYTE,S.A. 1975, Distribution, trophic relationships and breeding habits of the fish populations in a tropical lake basin (Lake Bosumtwi, Ghana) . J.Zool..Lon&n,l77,25-56.

. . . .

, . ..: