Quaternary history of the Lake Magadi Basin, southern Kenya Rift: Tectonic and climatic controls

Highlights

• Progressive increase in palaeolake alkalinity and salinity over the past million years.

• Middle Pleistocene to Holocene increase in aridity was interrupted by climate-induced wetter episodes.

• Diatom floras indicate episodically meromictic palaeolakes.

• Step-like changes in sedimentation reflect faulting and drainage diversion.

• Faulting tapped deep groundwater releasing Si-enriched deep fluids and mantle CO₂ to produce abundant chert and thick trona deposits.

Abstract

Sediments from the Magadi Basin (south Kenya Rift) preserve a one-million-year palaeoenvironmental record that reflects interactions between climatic, volcanic and tectonic controls. Climate changes that impacted sedimentation include wet-dry cycles on variable timescales and an overall progressive trend towards greater aridity. Volcanic influences involved inputs of tephra to the basin, significant inflow of geothermal fluids, and the effects of weathering, erosion and transportation of clastics from trachyte and basalt terrains. Tectonic controls, which were often step-like, reflect the influence of faults that provided pathways for fluids and which controlled accommodation space and drainage directions.

Intensified aridity and evaporative concentration resulted in salinity and pH increasing with time, which led to a change from calcite deposition in mildly saline lakes before 380 ka to the later formation of zeolites from reactions of volcaniclastic debris with highly alkaline lake and pore water. After 105 ka, hyperalkaline conditions led to trona accumulation and increasingly variable rare earth elements (REEs). The presence of mixed saline and freshwater diatom taxa between 545 and 16 ka indicates climate variability and episodic inputs of fresh water to saline lakes. Calcrete formed in lake marginal settings during semi-arid periods.

Tectonic controls operated independently of climate, but they interacted together to determine environmental conditions. Aquatic deposition was maintained during periods of increasing aridity because fault-controlled ambient and geothermal springs continued to flow lakewards. This recharge, in turn, limited pedogenesis: palaeosols are common in other rift floor sequences. Trona formed when aridity and evapoconcentration increased, but its precipitation also reflects increased magmatic CO₂ that ascended along faults. Basin fragmentation and north-south fractures caused loss of cross-rift (east-west) drainage from rift-marginal basalts, resulting in reduced transition metals after 545 ka. The Magadi Basin demonstrates how a careful reconstruction of these complex tectono-climatic interactions is essential for accurate palaeoenvironmental reconstruction in continental rifts and in other tectonic settings.

Introduction

Much of the early palaeoclimate research in East Africa was devoted to understanding Plio-Pleistocene hominin and archaeological sites (see Cohen et al., 2016 and Campisano et al., 2017 for reviews). More recently, efforts have focused on climate variability from various Pliocene to Recent time-slices (Kingston et al., 2007; Owen et al., 2008; Tierney et al., 2010; Junginger and Trauth, 2013; Magill et al., 2013). A range of record types have proven useful in reconstructing palaeoclimates in the region, including Late Quaternary core-records (e.g., Verschuren and Chapman, 2008; De Cort et al., 2013, De Cort et al., 2018), outcrop data (e.g., Levin, 2015), and marine cores (e.g., deMenocal, 1995, deMenocal, 2004). More recently, drill cores from extant rift lakes have proven exceptionally useful in documenting palaeoenvironmental histories, including a 1.3-million-year Lake Malawi record (Ivory et al., 2016). In addition, a one-million-year pollen-diatom-mineralogy study of a core at Lake Magadi synthesised the regional climatic history for the southern Kenya Rift as a basis for exploring hominin evolution and mammalian change (Owen et al., 2018a). The latter paper complements this study, which is based on outcrops and two lake cores, and which focuses on aquatic sedimentation and palaeoenvironments using sedimentological, mineralogical and geochemical data, supplemented by diatom analyses.

Lake Magadi lies in the axial trough of the southern Kenya Rift ~605 m above sea level (masl), and is a seasonally-flooded, saline alkaline pan, currently floored by trona (Fig. 1). Discontinuous Quaternary sedimentary outcrops around the lake include fluvial sediments (channel deposits, alluvium), calcrete, lacustrine limestone (including microbialites), zeolitic mudstone and siltstone, sodium silicate minerals, and chert of diverse origins (Baker, 1958, Baker, 1963; Eugster, 1967, Eugster, 1969, Eugster, 1980; Hay, 1968; Herrick, 1972; Surdam and Eugster, 1976; Behr, 2002; Brenna, 2016; Felske, 2016; Leet et al., 2016). These sediments accumulated in a N-S axial rift sump, in which the northern depocentre remained a lake or wetland for most of the last million years because of spring recharge during drier periods. Lake Magadi probably united with Lake Natron in northern Tanzania as a single, relatively dilute lake for different periods during the Pleistocene and early Holocene (Eugster, 1986; Casanova and Hillaire-Marcel, 1987; Williamson et al., 1993).

Lake Magadi was cored in June 2014 by the Hominin Sites and Paleolakes Drilling Project (HSPDP), which aims to develop basin-to-regional scale palaeoenvironmental histories that can be compared with local hominin remains and artefacts to infer possible environmental influences on hominin evolution (Cohen et al., 2016). We present here results of detailed analyses of the sedimentology, major- and trace-element geochemistry and mineralogy from two Lake Magadi cores and nearby sediment outcrops (Fig. 1A), supported by diatom records, which collectively provide a history of changing aquatic environments during the past million years. Both drill cores reached the volcanic basement. This new evidence gives an opportunity to reconstruct the Pleistocene history of the Magadi basin in much greater detail than previously possible. Specifically, we aim to: 1) reconstruct the environmental history of the Magadi palaeolakes; and 2) relate that sedimentary record to evolving tectonic, volcanic and climatic controls.