Humid glacials, arid interglacials? Critical thoughts on pedogenesis and paleoclimate based on multi-proxy analyses of the loess–paleosol sequence Crvenka, Northern Serbia

Abstract

Loess–paleosol sequences (LPS) provide unique terrestrial archives for past environmental and climate changes. The loess units are generally interpreted to reflect more arid conditions, whereas paleosols are thought to document more humid conditions. This paradigm may, however, deserve to be challenged in some regions, because it does not take into account past changes in dust accumulation rates. Here we present a multi-proxy sedimentological and geochemical characterization of the LPS Crvenka, Serbia, spanning the last ~140ka. Commonly used analyses, such as magnetic susceptibility, organic carbon content, grain size and weathering indices, allow the quantification of the intensity of pedogenesis. However, loess accumulation rates were much higher during glacials than during interglacials, so that it is not trivial to infer past rates of pedogenesis and related paleoclimate conditions. Novel proxies, based for example on lipid biomarkers and their isotopic composition, are not so directly affected by changes in loess accumulation rates and might provide interesting new insights.Long-chain, plant-derived n-alkanes in the LPS Crvenka suggest a minor, but non-negligible contribution of deciduous trees and shrubs (C27 and C29) during MIS 1, 2, 4 and 6, whereas grasses and herbs (C31 and C33) dominate the alkane input during MIS 3 and 5. This is in agreement with the notion that parts of southeast Europe served as refugia for deciduous trees during glacials. On the other hand, edaphic conditions on the drought-prone loess plateaus seem to have been more arid during MIS 3 and 5. Compound-specific δD analyses on the alkanes show only little changes on glacial–interglacial timescale. When compared with the isotopic enrichment of the Mediterranean Sea during the last glacial, this likely indicates a complex combination of reduced temperatures, increased rainfall, reduced evapotranspiration and more advection of North Atlantic moisture. Although disentangling these factors is difficult, interpretation in terms of humid glacials and more arid interglacials seems to be supported also by other novel lipid biomarkers, so-called GDGTs (glycerol dialkyl glycerol tetraethers).