Decoding geochemical signals of the Schwalbenberg Loess-Palaeosol-Sequences-A key to Upper Pleistocene ecosystem responses to climate changes in western Central Europe

Loess deposits are the most extent continental archives of climatic-and environmental change and represent important components of local and global dust systems. Consequently, their geochemistry provides an excellent basis for studying climate oscillations on land and how these affect processes in the terrestrial system. It is, however, challenging to assess information archived in loess records due to complexities in their formation, causing interfering geochemical signatures. In particular, the use of element ratios to derive weathering indices may be complicated since dust sources change through time, and since ecosystems respond differently to changing conditions. Whilst this complexity appears as limiting factor, the explanatory potential of proxies increases when the diversity of geochemical processes is acknowledged. To make use of this potential, we integrate the decadic logarithm of element ratios indicative for provenance shifts, mineral sorting during transport and sediment reworking as well as weathering into multivariate statistical analysis. We test, if the sensitivity of Principal Component Analyses (PCA) and Linear Discriminant Analysis (LDA) can be increased by applying both to sub-datasets of diagnostic Stratigraphic Units (SU) of the Schwalbenberg Loess-Palaeosol-Sequences (Middle Rhine valley, Germany). The selected site recorded sub-millennial Upper Pleistocene ecological changes in an unprecedented resolution for Central European LPS. Differences in operating functions of PCA and LDA and our age model highlight timing and intensity of provenance shifts, sediment relocation, decalcification, redox dynamic and clay-mineral formation associated with Atlantic-driven climate oscillations. Based on this we provide evidence for the accretionary character of interstadial palaeosols caused by a close interplay of dust input and soil formation. Automated re-detection of SU across the site using LDA functions contributes important insights into slope dynamics through regional erosion events. Overall, our approach provides a key for tracing ecological changes during climate oscillations across continents.