Holocene rainfall runoff in the central Ethiopian highlands and evolution of the River Nile drainage system as revealed from a sediment record from Lake Dendi

Highlights

• Well dated record of Holocene rainfall variability in the region of the Blue Nile headwaters

• Tephra deposition of a potential Mount Wenchi eruption at 10.2 cal kyr BP

• High variations in rainfall variability between 10.2 and 6.4 cal kyr BP

• Highest aridity around 3.9 kyr BP

• The shift of the ITCZ during the Holocene can be traced into the eastern Mediterranean.

Abstract

A 12 m long sediment sequence was recovered from the eastern Dendi Crater lake, located on the central Ethiopian Plateau and in the region of the Blue Nile headwaters. 24 AMS radiocarbon dates from bulk organic carbon samples indicate that the sediment sequence spans the last ca. 12 cal kyr BP. Sedimentological and geochemical data from the sediment sequence that were combined with initial diatom information show only moderate change in precipitation and catchment runoff during that period, probably due to the elevated location of the study region in the Ethiopian highlands. Less humid conditions prevailed during the Younger Dryas (YD). After the return to full humid conditions of the African Humid Period (AHP), a ~2 m thick tephra layer, probably originating from an eruption of the Wenchi crater 12 km to the west of the lake, was deposited at 10.2 cal kyr BP. Subsequently, single thin horizons of high clastic matter imply that short spells of dry conditions and significantly increased rainfall, respectively, superimpose the generally humid conditions. The end of the AHP is rather gradual and precedes relatively stable and less humid conditions around 3.9 cal kyr BP. Subsequently, slightly increasing catchment runoff led to sediment redeposition, increasing nutrient supply, and highest trophic states in the lake until 1.5 cal kyr BP. A highly variable increase in clastic matter indicates fluctuating and increasing catchment runoff over the last 1500 years. The data from Lake Dendi show, in concert with other records from the Nile catchment and the Eastern Mediterranean Sea (EMS), that the Blue Nile discharge was relatively high between ca. 10.0 and 8.7 cal kyr BP. Subsequent aridification peaked with some regional differences between ca. 4.0 and 2.6 cal kyr BP. Higher discharge in the Blue Nile hydraulic regime after 2.6 cal kyr BP is probably triggered by more local increase in rainfall, which is tentatively caused by a change in the influence of the Indian Ocean monsoon.

Introduction

The present day climate of Northeast Africa is affected by complex interactions between the West African Monsoon, the Intertropical Convergence Zone (ITCZ), and the Congo Air Boundary (CAB; e.g. Ivory et al., 2012; Costa et al., 2014; Enzel et al., 2015), where humid air from the Atlantic and Congo Basin converge with drier air masses from the Indian Ocean (see also Baker et al., 2007). Sediment records from the marine (e.g. Arz et al., 2003; Schmiedl et al., 2010; Ehrmann et al., 2013; Weldeab et al., 2014; Rohling et al., 2015) and terrestrial realm (e.g. Dramis et al., 2003; Fleitmann et al., 2003; Garcin et al., 2012; Foerster et al., 2012; Ghinassi et al., 2012; Costa et al., 2014; Junginger et al., 2014) provide information on the Late Pleistocene and Holocene climate in the region. These records show a broad agreement in the general climate pattern, but show also a number of spatial and temporal differences.

The existing paleoclimatic records indicate more humid conditions between ca. 15 and 5 cal kyr BP, a period known as the African Humid Period (AHP; e.g. deMenocal et al., 2000; Renssen et al., 2003; Tierney et al., 2011a; Junginger and Trauth, 2013; Shanahan et al., 2015). However, the exact timing, intensity, and abruptness of the onset and termination of the AHP are still under debate (e.g. Tierney and deMenocal, 2013; Costa et al., 2014). The same is true for short spells of relatively dry conditions in East and Northeast Africa, such as the Younger Dryas (YD), which interrupted the AHP for approximately 1000 years (e.g. Gasse et al., 1989; Johnson et al., 2000; Talbot and Lærdal, 2000; Tierney and deMenocal, 2013; Junginger et al., 2014). The termination of the AHP is rather abrupt in some records from Northeast Africa (e.g. Garcin et al., 2012; Tierney and deMenocal, 2013; Costa et al., 2014) while other records indicate a more gradual aridification at the end of the AHP (e.g. Kröpelin et al., 2008; Foerster et al., 2012; Junginger et al., 2014). Regional differences in hydrological conditions are also recorded during the late Holocene, with drier conditions during the Medieval Climate Anomaly (1250–950 CE; MCA) in the Ethiopian Arsi Mountain range (Bonnefille and Mohammed, 1994) and in central Kenya (Verschuren et al., 2000; Cort et al., 2013) and wetter conditions at Lake Tanganyika (Tierney et al., 2008, Tierney et al., 2010) and in northern Kenya (Brown and Fuller, 2008; Garcin et al., 2012; Junginger et al., 2014). Also during the Little Ice Age (1850–1550 CE; LIA), some proxy records of East and Southeast Africa (Nicholson et al., 2013) show a drier climate, whereas others infer increased humidity (e.g. Bonnefille and Mohammed, 1994; Verschuren et al., 2000; Darbyshire et al., 2003).

These local and spatial hydrological differences affect the drainage systems in eastern and northeastern Africa, including that from the Nile river. Increased freshwater supply from the Nile river (e.g. Emeis et al., 2000; Hennekam et al., 2014; Weldeab et al., 2014; Williams et al., 2015b) or Saharan (Rohling et al., 2002, Rohling et al., 2015; Osborne et al., 2008) and Sudanian (Woodward et al., 2015) wadi systems during the intensification of the African monsoon system (review in Rohling et al., 2015) contributed significantly to the formation of sapropel S1 in the eastern Mediterranean Sea (EMS) between 10.2 and 6.4 cal kyr BP. The associated suspension load from the Nile can be traced as far as into the South Aegean Sea (e.g. Ehrmann et al., 2007). Although the Blue Nile contributes at least 60% to the main Nile discharge, particularly during summer (Dumont, 1986; Rohling et al., 2015), hydrological records from its 176,000 km² large catchment (Conway, 2000) are sparse. Sediment records from Lake Tana, which is considered to be the headwater of the Blue Nile, provide important paleoenvironmental information over the last 17 cal kyr BP (Marshall et al., 2011; Costa et al., 2014), but the catchment of Lake Tana covers only 15,000 km² (Wale et al., 2009).

The aim of this paper is to better understand hydrological changes in the Blue Nile catchment using a sediment sequence from the eastern Dendi Crater lake on the central Ethiopian Plateau. Furthermore, the comparison of the records from Lake Dendi, from Lake Tana ca. 350 km to the NNE, from lakes that tentatively contribute to the White Nile hydraulic system, and from the EMS allows for a better discrimination between spatial and temporal climate patterns in the Blue and White Nile hydraulic systems and their influence on the sapropel S1 formation in the EMS, as well as a better understanding of large-scale changes in the atmospheric circulation patterns in the region.