Phytoliths and the Late Quaternary vegetation history of West Africa

点击翻译按钮获取中文摘要

Phytoliths (= plant silica bodies) are a valuable palaeoecological tool and complementary to pollen and plant macro-fossils. Phytolith studies on sediments for which other botanical proxies are available allow for combining and calibrating the different signals. While grass pollen is only identifiable to family level, with the highly diverse grass short cell phytoliths (GSCP) modern and ancient grasslands can be differentiated. This is especially important in Africa where savannas and grasslands cover the largest part of the continent. The major objective of the proposed project is an innovative contribution to the vegetation history of West Africa through the study of phytolith assemblages in modern plants and soils, and in Late Quaternary lake and fluvial sequences for which other botanical proxies (pollen, wood charcoal) have already been published. In West Africa, three major issues in palaeo-phytolith research merit attention and will be considered by the project: 1) distinguishing grassland communities, 2) assessment of the woody plant cover (forest-savanna question), 3) detecting human impact. The research is organized into three complementary work packages: WP 1) We will study modern soil phytolith assemblages along a precipitation gradient of 400-1300 mm from the Sahel to the Guinean zone. In Burkina Faso, the major objective is the separation of climatic, edaphic and anthropogenic signals in the phytolith spectra of different savanna vegetation communities. In Benin, we will study the soil phytoliths of gallery and semi-deciduous forests for a better discrimination of forest types. For characterizing the grassland, savanna and forest communities, we will use a recently developed GSCP classification system and detailed morphotypes characteristic for woody plants (see WP3). Commelinaceae phytoliths, first described by our working group, will help to detect and quantify human impact. Based on the statistical evaluation of the modern assemblages, we will test new phytolith indices to apply them, in a second step, to the fossil samples. WP 2) We will reconstruct the development of Holocene and Late Pleistocene grasslands and their correlation with climatic fluctuations for three well-dated palaeoecological sequences of the Sahelian and the Sudanian zone (Lake Tilla and Manga Grasslands / Nigeria; Ounjougou / Mali). We will also test if human impact, mainly agriculture, is detectable in the Late Holocene Sahelian and Sudanian assemblages, as suggested by archaeological evidence. For the sequence of Lac Sélé / Benin the major question is how the distinct shift from forest to savanna around 4500 cal BP, as visible in the pollen diagram, is represented in the phytolith record. WP 3) For a better interpretation of the signal for woody plants in the modern and ancient sediments, we will study herbarium material from modern trees and shrubs to identify potentially diagnostic morphotypes