Earliest fossil pollen records of endemic African Sclerosperma palms and the palaeoecological aspects of the genus.

Palm diversity in present-day Africa is much less than in other tropical regions such as the Americas and Southeast Asia but comprises several endemic genera, including Sclerosperma. The origin and evolution of this small Arecaceae genus are still enigmatic, and every new fossil discovered helps fill the void in its palaeophytogeographic history. In this article, we report five different pollen morphologies assigned to Sclerosperma from the earliest Late Eocene (early Priabonian) of southeast Kenya, Africa. These pollen grains represent the earliest fossil records of Sclerosperma so far and add a new geographic region and time slot to the palaeophytogeographic history of this genus. The morphological variability observed in the fossil pollen suggests that the diversification within Sclerosperma was already underway by the onset of the Priabonian, c. 37 Ma. Detailed analysis of the distribution of the three extant Sclerosperma species shows that this genus occurs only under Equatorial/Tropical climates with MTCM ≥ 18 °C, such as fully humid Rainforest (Af), Monsoonal (Am), and winter dry Savannah (Aw) climates. Sclerosperma is also conservative when it comes to vegetation biomes and only occurs in three different but closely related biomes, the Tropical & Subtropical Moist Forests Biome, the Tropical & Subtropical Grasslands, Savannahs & Shrublands Biome, and the Mangroves Biome. No current outliners pinpoint to a wider ecological tolerance of this genus in the geological past. This suggests Sclerosperma to be an exceptionally reliable taxon for palaeoclimate estimations and when assigning palaeofloras to major vegetation biomes. • Eocene Sclerosperma pollen grains are the oldest fossils from this enigmatic palm. • Pollen morphotypes suggest intrageneric divergence prior to the Late Eocene. • Fossil records of Sclerosperma suggests they were always endemic to tropical Africa. • Sclerosperma fossils can be used to estimate paleoclimate and past vegetation biomes. [ABSTRACT FROM AUTHOR]