PP31A-2211
Late Pleistocene-Holocene paleoclimate in southwestern Brazilian Amazonia with basis on floristic changes interpreted from isotope data

Wednesday, 16 December 2015
Poster Hall (Moscone South)
Dilce Fatima Rossetti, Organization Not Listed, Washington, DC, United States, Marcelo C.L. Cohen, Universidade Federal do Pará, Department of Geosciences, Belém, Brazil and Luiz C.R. Pessenda, CENA, Laboratory of 14C, Piracicaba, Brazil
Abstract:
Previous late Quaternary paleoclimatic interpretations in Amazonia have considered fluctuating dry to wet episodes with changes from savanna to forest, a view that concurs with other proposals of undisturbed rainforest despite global oscillations. Most of this debate is based on pollen data, but such elements are scarce in Amazonian sedimentary records. This work interprets vegetation in the Late Pleistocene and Holocene in a southwestern Amazonia lowland using δ13C, δ15N, C/N integrated with geomorphology, sedimentology and radiocarbon dating. The goal was to reconstruct vegetation changes through time and analyze their relation to climate and sedimentary dynamics. Fluvial channel and floodplain deposits with phytoplankton, as well as C3 and C4 land plants, were recorded. Between 42,033−43,168 cal yr BP and 34,804−35,584 cal yr BP, C4 land plants increased as a result of a climate drier than today´s. However, wet climate prevailed from this time-frame until the onset of the Last Glaciation Maximum. In the Pleistocene/Holocene transition, there was an increased contribution of C4 land plants potentially related to dry episodes. However, the increased contribution of this type of land plant is not synchronous with Holocene dry episodes previously documented for the Amazonian lowland. On the other hand, it is remarkable that the record of this plant type was verified only in sites with modern grassland confined to fluvial paleo-landforms. Thus, rather than due to a dry climatic episode, the recorded grassland expansion and its maintenance up to the present time in the studied sites is more likely associated with the evolution of depositional environments, being coincidental with the progressive abandonment of fluvial systems. An important conclusion derived from the present work is that great care must be placed when reconstructing late Quaternary paleoclimate in Amazonia based on changes in floristic patterns, as they may be also a response to sedimentary dynamics.