A method to determine warm and cool steppe biomes from pollen data; application to the Mediterranean and Kazakhstan regions

An objective method for the assignment of pollen spectra to appropriate biomes has been published recently. the aim of this paper is to improve the distinction between warm and cool steppes, thus refining vegetation and climate reconstruction, particularly during the Last Glacial Maximum. A set of modern pollen spectra from the Mediterranean and Kazakhstan regions, dominated today by open vegetation types, has been analysed statistically in order to relate pollen taxa abundances to warm and cool grass/shrub plant functional types (PFTs). A statistical test using modern pollen data shows that the method is able to distinguish between cool and warm steppe biomes with a high degree of confidence. the method has been applied to two tossil pollen records. the results of this exercise showed that cool steppe dominated in central Greece between 18 000 and 13 000 yr BP, while in western Iran the vegetation was at the boundary between cool and warm steppes. these vegetation types were replaced by warm mixed forest in Greece and warm steppe in Iran after that time span.