Charles Explorer logo
🇬🇧

A Landscape Reconstruction Algorithm and pedoanthracological data reveal Late Holocene woodland history in the lowlands of the NE Czech Republic

Publication at Faculty of Science |
2017

Abstract

We aim to obtain composition of the regional vegetation from the pollen record in the high mountains, to use it in the interpretation of the pollen record from the lowlands via a Landscape Reconstruction Algorithm (LRA), and to then compare the pedoanthracological data with the LRA result based on dissimilarity coefficient. We used five pollen sequences from summits of the Eastern Sudetes (NE Czech Republic), and two pollen cores were analysed in the adjacent Litovelske Pomoravl lowlands.

The same lowland woodland was sampled in six pedoanthracological sections. All records cover the Late Holocene period.

Site-to-site variability of the pollen spectra from the mountains was found to be low. Regional vegetation inferred from shallowest pollen samples corresponds to present-day vegetation inferred from forestry data.

Mountain sites represent regional components and can substitute for large sites in an LRA. Mean LRA estimates were more similar to mean charcoal spectra than they were to mean pollen percentages.

LRA estimates of Quercus and Fraxinus in the lowlands match the pedoanthracological evidence regarding individual sites. LRA estimates and charcoal spectra both show that the main dominant (>50%) in the woodland studied was Quercus.

Mismatching evidence resided in a small number of the determined charcoals and there were differences between some assumptions of the LRA and the real-world conditions. Alnus and Tilia produced higher pollen signals on the expanse of Fagus and Abies, which were recorded by charcoals within the Relevant Source Area of Pollen (RSAP), which varied from a 35 to a 255 meter radius, and ranged with the value from forest hollows.

Comparability of LRA vegetation estimates and charcoal percentages is still at the level of complementary evidence. However, a quantitative approach shifts the character to substitutable evidence, with potential to provide the same vegetation estimates by both proxies.