Holocene variations of river drainage in Libya

We analyzed Sr and Nd isotopes in deep-sea sediments cored in the Gulf of Syrte, Mediterranean Sea. Samples span the last 19 ka (dated by 14C) and include one prominent sapropel, S1. It had been proposed (e.g. Hoelzmann et al, Glob. Biogeoch. Cycles 12 (1998) 35-51, and references therein) that during the Early-Mid Holocene wet phase (ca. 9 - 5 ka BP) the river drainage network into the Mediterranean Sea was greatly expanded and reached the Tibesti Massif. In that case, we expect variations in the Sr and Nd isotopic records, due to differences in the eroded basement rocks. Our results demonstrate a good correlation between Sr and Nd isotopic systems. Epsilon Nd values, eps(Nd), for present-day sediments and sediments younger than 2.5 ka are uniform between -12.3 and -11.8. Eps(Nd) increases down-core, culminating at -9.8 at 6.5 ka BP, then decrease again and reach -11.0-11.5 during the glacial dry period, similar to, but resolvably different from, the present-day. The Nd model ages vary slightly from 1.8 to 1.7 Ga; while the difference may be not significant, their trend is quite clearly parallel to the eps(Nd) trend. The Sr isotopic ratios range between 0.70966 and 0.70934, in anticorrelation with eps(Nd). At present day, low eps(Nd) is explained by drainage basins being limited to the coastal areas and by eolian sand input, both reflecting continental crust mostly no younger than Paleozoic. During the 5-9 ka BP wet phase, erosion of the young, mafic Tibesti volcanics shifted sediment signatures to higher eps(Nd) and lower Sr isotopic ratios. During the glacial maximum, long-distance sediment transport from Tibesti to the sea was as low as during the last 1.5 ka, at the height of Sahara desertification. The observed small differences in Nd and Sr between present-day and the glacial maximum may indicate that eolian dust transport was less important at that time. We conclude that the river drainage in Libya reached farther south during the Early-Mid Holocene wet phase.