Guillaume Buchet

Palynology, stratigraphy and palaeoenvironment of a pliocene hominid site (2.9-3.3 M.Y.) at Hadar, Ethiopia

Abstract The Pliocene Hadar Formation (2.9-3.3 m.y.) is well known for its abundant fossil fauna and hominid remains. Fossil pollen data obtained from 27 distinct stratigraphic layers are presented together with tectonic and sedimentary data. Comparison with modern pollen data from the Awash basin (67 surface and fluviatile samples) indicates that montane evergreen bushland, presently known from above 1500 m, occupied the Hadar basin (500 m) and was associated with a humid montane forest. A downfaulting of about 1000 m presumably happened after 2.9 m.y. whereas a shift toward more arid grasslands occurred in the middle of the Hadar Formation.

Pollen-based vegetation changes in southern Tanzania during the last 4200 years: climate change and/or human impact

The age-constrained pollen data of a sedimentary sequence from the crater Lake Masoko, southern Tanzania (9‡20PS, 33‡45PE, 770 m), display a continuous record of vegetation for the past 4200 years. This record provides evidence that wetter Zambezian woodlands always occupied this area during the late Holocene, reaching a maximum extent between 2800 and 1650 cal yr BP related to increase in summer monsoon intensity. However, three main episodes of decline have been detected, between 3450 and 2800 cal yr BP, between 1650 and 1450 cal yr BP and from 1200 to 500 cal yr BP, for which a climatic interpretation, decrease in the summer monsoon strength, was preferentially advanced. The first is synchronous with lowstand of many tropical African lakes and, so, mainly induced by increased aridity. In contrast, the abrupt change in the pollen record at 1650^1550 cal yr BP is marked by a large extension of grasslands at the expense of arboreal cover, further by an increase in Ricinus communis and an intensification of burning. It could thus indicate local clearance of vegetation by man. However, at the same time, the decline of montane forest suggests the impact of a more regional change. During the last episode, between 1200 and 500 cal yr BP, dry climatic conditions are inferred from a combination of pollen, diatom and magnetic proxies, although the occurrence of Late Iron Age settlements in the region means that local human interference cannot be excluded. This study illustrates the difficulties in deciphering ecological and anthropological changes from pollen data in African tropical regions. A 2003 Elsevier B.V. All rights reserved.

Glacialanterglacial record from intertropical Africa, high resolution pollen and carbon data at Rusaka, Burundi

Abstract In northern intertropical Africa, many geological data have documented several abrupt climatic fluctuations taking place during the time of the last deglaciation. One of them is an aridity event, generally placed in the interval 11,000–10,000 BP, and synchronous with the Younger Dryas cold episode of the temperate regions, that has been explained by the influence of the Atlantic Ocean circulation on global climate. In this paper we present high resolution (50 to 150 years) studies of organic matter and pollen undertaken on a new core from the Rusaka peat bog (3°26′ South, 29°37′ East, 2070 m altitude), dated by the conventional (l2 dates) and the AMS methods (18 dates obtained on bulk sediment and macroremains). Although the sediment at the bottom of the core was deposited during the last glacial period prior to 21,000 BP, the dating provides a reliable chronology for the last 12,000 BP only. Several vegetational changes are documented in the interval 12,000–9000 BP, but their chronological placement can only be achieved with a few years of uncertainties. A great dispersion of AMS dates is attributed to contamination by charcoal dated between 11,700 and 11,500 BP, and to a possible age plateau at ca. 10,000 BP. Therefore the results are preferably presented versus depth to keep synchronism in the stratigraphic placement of the successive events. At Rusaka, increase in carbon storage, C/N ratio for terrestrial plants, and tree cover began at ca. 12,000 BP, soon after the swamp had been established. This date is in good agreement with a major increased hydrological balance well-documented throughout the intertropical region. The forest installation is progressive, rather complex and seen as a two step process. The first forest phase (11,700 BP to 10,600 BP) is dominated by Hagenia, a montane taxon, in good correspondance with previous records from other authors in East African mountains. The second forest regeneration took place post 10.000 BP, after a clearly reversed trend of tree cover from 10,600 to 10,000 BP, ending with a short episode of minimum trees lasting no more than about 200 years. But there is no indication of cooling whatsoever, and the opening of the forest is better explained by increased aridity, rather than normal pattern of forest succession. This will remain a controversal interpretation among palynologists until high resolution pollen data can be obtained for the 13,000 BP and earlier period. Nevertheless, the timing of the arid episode at Rusaka is in good agreement with other evidence mentioned from highland and lowland pollen data. A detailed comparison with 25 available pollen diagrams is limited by poor or inconsistent dating or low resolution of the pollen data. Nevertheless pollen from four of them document an arid event in good agreement with its timing at Rusaka and high resolution isotopic results from two sites in equatorial lowlands and from the Sahel. This work shows that the arid Younger Dryas occurred at the same time both in lowlands and highlands, and therefore was most likely a precipitation signal rather than a response to temperature decrease.

Pollen-rain-vegetation relationships along a forest-savanna transect in southeastern Cameroon.

Modern soil and litter samples from southeastern Cameroon, collected along a continuous forest-savanna transect were analysed for pollen content to define modern pollen-vegetation relationships. The pollen results, completed and compared with botanical inventories, leaf area index and basal area measurements performed in the same area, clearly registered the physiognomy, the main floristic composition and floral richness of the two sampled ecosystems. Distortions were observed between sampled vegetations and their pollen rain, related to important differences in pollen production and dispersal of plant species: this is a general feature in many tropical regions. The pollen data in the area studied reflected well the recent transgression of forest versus savanna. This permitted us to define inside the forest ecosystem more successional vegetation communities than the botanical surveys allowed.

Nouvelle séquence pollinique d'une tourbière de la crête Zaïre-Nil (Burundi)

This paper presents the pollen analysis of a core, 11.5 m long, from a valley peatbog in the Burundi highlands, east of Lake Tanganyika. Nine radiocarbon dates provide a chronological frame of ca. > 32,500 to 3200 yr. B.P. At the bottom of the fossil sequence, pollen spectra indicate that open grassland occurred at 2000 m altitude during the last Glacial period. The pollen diagram mostly documents the composition and history of the humid montane forest during the Holocene time. The forest starts at ca. 9500 B.P. extrapolated date, with maximum development maintained until 6000 yr. B.P. Several forest decreases are documented since 4000 yr B.P., without any clear pollen indicator of anthropogenic activity. The most important one at ca. 3200 yr B.P., appears synchronous with an early date for the Iron Age in Burundi.