Abstract Profiles of the abundance or mass accumulation rate of biogenic silica in sediment cores can provide important insight into the past productivity of diatoms and the climatic influence on a lake basin through time. While primary production of diatoms in the overlying waters can influence the burial rate of diatoms, other factors need to be considered as well. The geochemical mass balance of dissolved silica in the lake system is one such factor that plays an important role in controlling the rate at which diatoms can be buried in the lake sediments. Examples are provided from Lake Malawi and Lake Victoria, East Africa, that demonstrate how the rates of input and outflow of dissolved silica are related to biogenicsilica profiles in the sediments, and how conclusions were drawn on past environmental conditions in these large tropical lakes.

Abstract Deposits from the modern highstand lacustrine deltas in Lake Malawi offer an excellent opportunity to test models of deltaic sedimentation in a tectonically active setting and to examine variations in sand-body geometry along the axis of a large lake that has significant gradients in physical processes. In 1991 we initiated a coring project in five of the largest deltas in the lake to describe, for the first time, the shallow-water environments of deposition and processes of sedimentation. Percussion drill holes through the lower delta plain in the Linthipe and Dwangwa (shoaling margin) deltas revealed moderate to extreme lithologic variability with sand units up to 15 m thick. Deposited in an alluvial or shallow subaqueous deltaic setting, the sediments ranged from clay to gravel, dark green to brown to orange in color, and contained sections with significant amounts of organic material. Units of gravel up to 1 cm in diameter were recovered, and 60 to 70 percent of the recovered sediments consisted of at least 75 percent sand and gravel. Silts and clays occurred in units up to about 1 m thick, usually in the middle section of each sequence. Although the age, and thus the sediment accumulation rate, is unknown, it is likely that the cored sections (20-26 m deep) represent a few hundred to a few thousand years of accumulation.

Abstract High-resolution seismic profiles, side-scan sonar records, and sediment cores collected from Lake Malawi have been analyzed to determine the nature of sedimentation in a modern rift lake. More than 4500 m of sediment have accumulated in the deepest basin in the northernmost part of the lake. If the modern sedimentation rate of 1 mm/yr is representative of most of the lake's history, then the deepest basin may, when compaction is accounted for, be on the order of 26 Ma. Although it is an open-basin lake at present, it has several times in the past been a closed-basin lake in response to drier climate. Lake level has been 100-150 m lower than present at least three times in the last 10 k.y. The distribution of modern sediment is quite complex. Little or no deposition occurs in most regions shallower than 100 m due to storm-generated surface-wave activity. Gravitational transport of sediment by creep, debris flows, slumping, and turbidity currents is common, particularly off deltas and border faults. Diatom-rich clays occur in the deep basins far removed from major terrigenous input. These typically have organic carbon concentrations of 3-6 wt. %. Laminated sediments are common in Lake Malawi, and their frequency increases with water depth. Neither the abundance of organic carbon nor frequency of laminations increases abruptly below the depth of the chemocline. This probably results from the rise and fall of the chemocline as well as lake level in response to climatic changes. Although source rock potential of Lake Malawi as a future petroleum resource is high, the reservoir rock potential has yet to be demonstrated.

The thickness of the unconsolidated sediment and the topography of the underlying bedrock surface of Lake Superior are interpreted from 8,000 km of high-resolution seismic reflection profiles taken during 1966 and 1967. A depth-to-bedrock map was constructed by combining the isopach map of unconsolidated sediments (prepared from our profiles) with the bathymetric map of the lake (Canadian Hydrographic Service Chart 885, 1973). Lake Superior can be divided into three morphologic regions on the basis of bathymetry and underlying bedrock: a western region composed of long linear valleys and gentle changes in relief, a central region composed of a single broad bathymetric depression, and an eastern region composed of a complex pattern of linear troughs and ridges. The western region is dominated by a nearly continuous bathymetric and bedrock valley paralleling the north shore from Thunder Bay, Ontario, to Duluth, Minnesota. The bedrock valley reaches depths of more than 800 m below lake level near Silver Bay, Minnesota, and has more than 500 m of overlying unconsolidated sediments. It probably resulted from differential glacial erosion, where the relatively erodible sandstone of the Proterozoic Bayfield Group comes in contact with Proterozoic rocks of the relatively resistant underlying volcanic rocks of the Keweenawan Supergroup and gabbro of the Duluth Complex. The central region, separated from the western region by a north-south basement ridge, consists of a broad valley with only 8 to 15 m of unconsolidated sediments overlying the bedrock surface. The complex pattern of troughs and ridges of the eastern region forms a north-south dendritic pattern with valleys as much as 100 km long but only 5 to 10 km wide. The bedrock surface is more than 600 m below lake level in some places and is overlain by as much as 300 m of unconsolidated sediments. These valleys are probably the result of erosion by subglacial streams. The stream erosion may well have followed a system of shear zones that have been observed in the shore exposures of the underlying Proterozoic sedimentary rocks of the Bayfield Group and Jacobsville Sandstone to the south. Most of these bedrock valleys are truncated about 15 km north of the south shore, where the lake floor rises abruptly to the coastline, although two valleys extend onshore. In general, the morphology of the bedrock surface in Lake Superior probably reflects the result of scour by glacial and subglacial streams, which in turn were localized by lithologic contacts, pre-existing topography, and shear zones. Over most of the lake, the acoustic impedance contrast across the contact of unconsolidated sediments with bedrock is high enough so that the seismic energy is reflected with little or no penetration into the bedrock. Despite this, there are some places where layering within the bedrock can be identified and apparent dips determined. Among the structural trends determined from these dips are the following: a southwest-plunging synclinal feature bordering the Bayfield Peninsula; a syncline lying between the Apostle Islands and the Keweenaw Peninsula that probably represents the center of the Lake Superior depositional syncline; a south-plunging syncline located between Michipicoten Island and Superior Shoals; and an apparent southward dip of the bedrock in the southeastern region of the lake.