Estimation of late Eocene (Bartonian–Priabonian) terrestrial palaeoclimate: Contributions from megafloral assemblages from central Germany

• We reconstructed terrestrial palaeoclimate of late Eocene plant assemblages.
• The palaeoclimate estimates rest upon quantitative approaches (CLAMP, LMA, CA).
• The data indicate similar palaeoclimate conditions as predicted by marine proxies.

The Eocene, the world’s last global greenhouse period, shows distinct climatic developments through time, from subtropical towards warm-temperate conditions in Europe. The present paper investigates the response of five floras of the central German Weißelster Basin from different stratigraphic ages to terrestrial palaeoclimatic variations. The floras were selected from the Zeitz Floral Assemblage (“Florenkomplex”) and cover a time interval of ca. 3.5 Myr within the late Eocene, which is supposed to be a period of gradual decline of global mean annual temperatures. The Zeitz Floral Assemblage was characterized by mainly evergreen, notophyllous intrazonal vegetation, consisting of subtropical and warm-temperate elements. The five floras as well as the entire Zeitz Floral Assemblage have been analysed using the following quantitative techniques: leaf margin analysis (LMA), Climate Leaf Analysis Multivariate Program (CLAMP) and Coexistence Approach (CA). Furthermore, the results have been compared with previously published estimates for additional localities from the same area and age. The temperature estimates derived for the Zeitz Floral Assemblage are in a good agreement with those of previous studies i.e. mean annual temperature (MAT) 16.1–20.0 °C, warmest month mean temperature (WMMT) 24.9–26.4 °C, coldest month mean temperature (CMMT) 9.2–13.3 °C. Deviations have been determined for the calculated mean annual precipitation (MAP 1308–1335 mm) compared to previously published data, indicating values of about 2000 mm. Both CLAMP and CA reveal clear hints for seasonality in precipitation which is in agreement with previous phytosociological assumptions and palynological data. The estimates for the individual sites differ between the different techniques as well as among the individual sites, but generally indicate no temporal trends in temperature and precipitation development. Additionally, the applicability of these techniques for fluvial-estuarine plant assemblages, as well as the applicability of intrazonal vegetation for reconstructing transregional climate conditions are discussed. Taphonomic and methodical biases are also subject of debate.