Timing and stepwise transitions of the African Humid Period from geochemical proxies in the Nile deep-sea fan sediments

Highlights

• Reliability, with qualifications, of brGDGTs as terrigenous flux tracers in deltaic environments.

• Diversity of sedimentary settings in the Nile deep-sea fan.

• Fluctuations of terrigenous inputs as retraced by εNd and brGDGTs records driven by summer insolation.

• Stepwise onset of the AHP at 14.8 ka and early initiation of the AHP termination at 8.5 ka.

• Stepwise termination pattern to a modern hydrology state occurred within 4 ka.

• Imprint of abrupt “North Atlantic events” such as the Younger Dryas in the Nile sedimentary sequence.

Abstract

Large fluvial systems, such as the Nile River, allow a basin-scale integrated view of climatological and environmental changes. In this study, we reconstructed the Nile discharge history for the last 20 ka using molecular ratios of glycerol dialkyl glycerol tetraethers (GDGTs) and neodymium (Nd) radiogenic isotopes. By characterizing both the organic and inorganic fractions, we assessed the relevance of the GDGT-based proxies in deltaic environments as tracers of terrigenous origin. A large increase in Nile discharge is documented from 14.8 to 8.4 ka BP reflecting enhanced physical erosion and transport processes from the Ethiopian Traps. We confirmed the primary control of insolation on precipitation on North East Africa through the last 20 ka. The centennial time resolution reached on the sedimentary sequence revealed a step-wise onset and termination of the African Humid Period (AHP) starting at 14.8 ka and ending at 8.4 ka BP, respectively. Our centennial-millennial records allowed furthermore pinpointing the abrupt periods of arid conditions corresponding to the Younger Dryas. These data illustrate the linkage between low and high latitude hydrological variability.

Introduction

African climate varied greatly during the late Quaternary with alternating periods of aridity and humidity (Singarayer and Burrough, 2015). The last major humid phase in Northern Africa, i.e., the African Humid Period (AHP, deMenocal et al., 2000; Shanahan et al., 2015; Tierney et al., 2011), occurred during the early and mid-Holocene (∼11–5 ka BP, Gasse, 2000). The development of the so-called “Green Sahara” during the AHP allowed the establishment of large lake systems in Northern Africa and vegetation in areas that are now desert. Human colonization was possible at latitudes 800 km further north than now (Kropelin et al., 2008; Kuper and Kropelin, 2006). This wet African phase has been attributed to the northward migration of the rain belt associated with the Intertropical Convergence Zone (ITCZ) induced by precession-forced insolation changes. These changes started at the end of the Last Glacial Maximum (LGM, Caley et al., 2011; Tuenter et al., 2003). The magnitude of this wet phase was also exemplified by the deposition of the most recent sapropel (S1) in the eastern Mediterranean Sea. Enhanced Nile river runoff led to surface water buoyancy gain, resulting in reduced deep-water ventilation and oxygen deficiency at depth (see Rohling et al., 2015; Rossignol-Strick et al., 1982; Vadsaria et al., 2019).

The termination of the AHP remains highly controversial, albeit frequently studied at many locations in and around the African continent (e.g., Collins et al., 2017; Shanahan et al., 2015; Tierney and deMenocal, 2013). Some studies suggest an abrupt transition that occurred at ∼5 ka BP (Collins et al., 2017; deMenocal et al., 2000; McGee et al., 2013; Tierney and deMenocal, 2013). Other studies propose that a gradual AHP termination (Berke et al., 2012; Foerster et al., 2012) or a termination that involved several phases of decreasing humidity (Liu et al., 2017; Loomis et al., 2015). In contrast, the onset of AHP is less controversial and began on the African continent around 15–14 ka (Shanahan et al., 2015; Trauth et al., 2018). However, within the Nile delta, and within the Eastern Mediterranean Sea, there are relatively few sites with a sufficient time resolution covering the last 20 ka. Thus, the onset of the AHP is less well described than its termination. On the Pleistocene time scale, humid periods are characterized by high sedimentation rates and a large proportion of smectite and Fe-rich silicate minerals within the Nile delta sediment (Caley et al., 2011; Emeis et al., 2000; Kroon et al., 1998; Revel et al., 2010; Zhao et al., 2011). These characteristics reflect increased physical erosion and material transport mainly originating from the Ethiopian highland during higher rainfall regimes related to summer monsoon precessional variability. Thus, the sediment records preserved at the Nile deep-sea fan provide a suitable archive to study basin-wide environmental changes in vegetation cover, lithology, and soil erosion/weathering at a high time resolution mainly during humid periods such as the AHP (Blanchet et al., 2014; Castañeda et al., 2016; Hamann et al., 2009; Hennekam et al., 2015; Weldeab et al., 2014).

Sedimentary provenance of terrigenous material in the Nile deep-sea fan can be characterized through its neodymium isotopic signature (εNd, Bastian et al., 2017; Blanchet et al., 2014; Blanchet, 2019; Revel et al., 2015; Weldeab et al., 2002). No significant fractionation occurs during chemical weathering so εNd reflects the isotopic composition of the parent bedrock and the origin of the material (Bayon et al., 2015). Since the Nile basin encompasses the Precambrian African basement as well as the Ethiopian and Somalian Cenozoic basalts with contrasting Nd isotopic compositions, geochemical studies on its sediment loads can distinguish between the sources of the material (Fig. 1A).

In parallel, glycerol dialkyl glycerol tetraethers (GDGTs), core membrane lipids of Achaea and Bacteria, are ubiquitous in marine to lacustrine waters and soils (e.g., Karner et al., 2001; Keough et al., 2003; Weijers et al., 2006). Specifically, the branched to isoprenoid tetraether ratio (BIT-index, Hopmans et al., 2004) and its first order derivatives are effective as tracers of continental inputs in estuarine settings (Blanchet et al., 2014; Kim et al., 2015; Ménot et al., 2006; Soulet et al., 2013). However, their environmental significance is controversial (French et al., 2015; Kaiser et al., 2015; Zell et al., 2015). The BIT-index depends on the fluctuations of branched GDGTs (brGDGTs) and variation of the GDGT isoprenoid crenarchaeol (Hopmans et al., 2004). The index is therefore dependent on processes that affect these GDGTs. BIT values are not just dependent on the inputs of soil brGDGTs, but also on the levels of marine Archaea. For example, sites with equal inputs of terrestrial brGDGTs but different local marine productivity can display different BIT values (Fietz et al., 2011; Herfort et al., 2006; Smith et al., 2010). Furthermore, differential degradation in the water column may affect the relative preservation of different compounds. Field and laboratory experiments have tested the stability of brGDGTs from soils to river mouths (Peterse et al., 2015; Zhu et al., 2011). Exposure of anoxic sediments to oxygen produced little degradation of brGDGTs. This was likely due to their enhanced protection by the sedimentary matrix inhered in soils. However, this protection might differ for water-column produced isoprenoid GDGTs (Huguet et al., 2008; Kim et al., 2009). To avoid these possible biases, the abundance ratio of the hexa-to penta-methylated brGDGTs ratio (IIIa/IIa) was proposed (Xiao et al., 2016) and its modified form ΣIIIa/ΣIIa (Martin et al., 2019). In this study we used the ΣIIIa/ΣIIa ratio along with the proportions of tetra-, penta- and hexa-methylated brGDGTs, and the #rings_tetra index defined as the weighted mean number of pentacycles in tetramethylated brGDGTs (Sinninghe Damsté, 2016).

Using inorganic (Nd isotopes) and organic (GDGTs) proxies, we characterized the erosion products in the Nile delta to assess past hydrological changes in its basin. We validated the reliability of the targeted sedimentary archive by testing our toolbox on the termination of the AHP. We then discuss the timing and pattern of this wet period as recorded in the Nile delta in GDGT-based indices and clastic εNd signatures.