Interpreting environmental signals from Pleistocene fluvial microvertebrates in the Middle Dnipro (Ukraine): Comparing habitat-weighting and bioclimatic-spectra approaches
Year: 2026
Type: preprint
Abstract: Fluvial microvertebrate assemblages represent an important but underexploited archive for reconstructing the Quaternary environments, particularly in the regions lacking karstic deposits. Here we analyse the late Middle and early Late Pleistocene small-vertebrate assemblages from the Middle Dnipro region (Ukraine) to evaluate the relative effectiveness of two widely used interpretative frameworks: habitat-weighting analysis and bioclimatic spectra. The studied assemblages span multiple Marine Isotope Stages (MIS 11 to MIS 5), deriving exclusively from alluvial contexts characterised by taphonomic overprinting and time-averaging. Using principal component analysis, we compare the environmental signals extracted by both methods across different taxonomic datasets. In both approaches, the dominant source of variation reflects changes in ecological evenness, expressed as shifts in the relative importance of steppe, tundra-steppe and high-latitude taxa. The most even assemblages correspond to cold-stage conditions and are characterised by the appearance of arctic and boreal elements, whereas low-evenness assemblages are dominated by core tundra-steppe taxa. Habitat spectra additionally capture geomorphological influences related to river-valley dynamics, resulting in a diffuse climatic signal. In contrast, bioclimatic spectra more directly resolve glacial-interglacial climatic gradients and show greater robustness to incomplete faunal lists and taphonomical biases. Our results demonstrate that bioclimatic spectra provide a clearer and more reliable palaeoclimatic signal than habitat-based approaches for fluvial microvertebrate assemblages. This method is therefore particularly well suited for the Quaternary environmental reconstructions in lowland river systems affected by strong taphonomic and sedimentary filtering.
Source: PaleoRxiv (OSF Preprints)
Language: English
Subfield: Atmospheric Science
Field: Earth and Planetary Sciences
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7223184292