Human impact overprinting the climatic signal in the last century of Lake Malawi's sediment record (East Africa)?
by
Maria Letizia Filippi
Centro di Studio per il Quaternario e l'Evoluzione Ambientale,CNR, c/o Dip. Scienze della Terra, Univ. La Sapienza, Ple aldo Moro 5, 00185 Roma, Italy.
Coauthors: Michael R. Talbot (Geological Institute, University of Bergen, Allégt. 41, 5007 Bergen, Norway)

We present a geochemical record of the last 700 yr from Lake Malawi (East Africa), the most southerly and second largest (by volume) of the East African lakes. The country of Malawi has seen its population more than double during the last century. Having a relatively small surface area and being characterised by a very hilly morphology, the population tends to concentrate in a few flat regions and along the lakeshore. Agriculture and fisheries represent the sole sources of income to the country, making it particularly dependent upon the health of the lake.

In 1998 an IDEAL expedition to the northern basin of the lake recovered five multi-cores (ca 0.5 m long) and six piston cores with companion trigger cores (8-9 m and 0.5-1 m long respectively) from five different location and depths. These triads of piston core, trigger core and multicore from the same site guarantee retrieval of the complete sequence, including the sediment-water interface. The total core record, two thirds of which is varved, covers the last 21 ky of Lake Malawi's history. We focus here on the upper varved section, characterised by beautiful couplets of diatom-rich and clay-organic-rich laminae, that reflect the seasonal change from the windy cool months (May to October) to the wet and warm months (November-April). This sequence is interbedded with two thin volcanic ashes and a few homogeneous clay layers with an erosive basal contact (homogenites) that allow correlation between cores. However, erosion in association with emplacement of the homogenites removed some of the varves, decreasing their potential temporal resolution. A chronology has been established by means of ²¹⁰Pb assays in the top 20-cm. and varve counts, allowing a resolution of ca 5 yr in the topmost part, and 20 yr in the rest of the upper section.

Two triads were selected for geochemical analysis (delta¹³C and delta¹⁵N of organic matter, TOC, Total Nitrogen content) with a sampling interval of 2 cm in the multicores, 4 to 10 cm in the trigger-cores, and 10 cm in the piston cores. Results from the composite record of the two triads are reasonably parallel, particularly in the topmost part (the last 300 yr). The most striking change is a shift of up to -2 permil in delta¹³C values of organic matter, starting in ca. 1880, which is paralleled by increases in TOC and Ntot, and a two- to three-fold increase in sedimentation rate. The only comparable delta¹³C shift of this magnitude is at the LGM transition, a change that seems to be linked to a major transgressive event, but spread over a much longer time interval. The ca 5 m transgression of Malawi during last century seems insufficient to alone explain the rapidity and intensity of this geochemical event. We suggest that these changes are more likely to be linked to the dramatically augmented human presence around the lake. Deforestation and land clearance for agriculture must have led to increased input of terrestrial organic matter, nutrients and detrital material to the lake. The role of possible changes in the rainfall regime during this period are thus difficult to assess, and human pressure on the lake system may be hiding any changes due to natural climatic variability.

Date received: April 3, 2001

Copyright © 2001 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cagc-97.