Characterization of neotropical forest and savannah ecosystems by their modern pollen spectra

Controversy surrounds the Quaternary palaeoenvironmental history of Amazonia. It is unclear whether moist evergreen forest, savannah or seasonally dry forest dominated the Amazon basin at the last glacial maximum (c. 21,000 years before present). The main source of information on past vegetation change in Neotropics are fossil pollen records collected from within ancient sediments. In part the controversy surrounding the palaeoenvironmental history of Amazonia stems from a poor understanding of these fossil pollen records. In order to improve interpretations of the fossil pollen record it is essential to better understand the nature of the pollen rain produced by modern ecosystems. Once the characterise pollen spectrum from an ecosystem has been established, it can be sought in the fossil pollen record. However, few such modern pollen-vegetation studies exist from the Neotropics and no one covers all the ecosystems that have controversial palaeoenvironmental histories.;In this thesis, the modern pollen rain from the three Neotropical ecosystems with controversial palaeoenvironmental histories are characterised. This is achieved by studying modern pollen rain from the Noel Kempff Mercado National Park area, Bolivia. to identify the characteristic pollen from these ecosystems four factors are considered: i) the spatial variation in the pollen rain, ii) the temporal variations in the pollen rain, iii) the statistically significant pollen types, and iv) the pollen-vegetation relationship. These analyses have revealed a small number of taxa that can be used to characterize these ecosystems. Further analyses have shown that the relative proportions of Didymopanax, Alchornea, Anadenanthera, Melastomataceae/Combretaceae, Moraceae/Urticaceae, Myrtaceae, Palmae, Pteropsidia (trilete), Poaceae and Solanum can be used to differentiate between these ecosystems. These findings have implications for the competing theories of Neotropical vegetation change. In the light of this research a comprehensive reassessment of the fossil pollen records from Amazonia is required