The past peoples of Amazonia: assessing ecological legacies


Amazonian forests harbor much of Earth’s biodiversity, yet the relative influence of processes driving spatial patterns of species richness and abundance remain unknown. An emerging hypothesis is that modern vegetation is largely a legacy of forest clearing and cultivation by ancient people, contradicting ideas of Amazonia as a ‘pristine wilderness’. Fossil pollen data are typically used to reconstruct past ecological states and identify past human activity, but to date no quantification of human influence has been attempted. Fossil pollen data are also used to quantify past climates using community-climate models (CCMs); however, observed and estimated climates are often mismatched. We hypothesize that the mismatch is related to human influence, particularly fire and forest clearing. To test this idea, we will produce a modern calibration dataset of: 1) observed climates, 2) observed human influence derived from satellite imagery, and 3) CCMs from modern pollen assemblages collected from the surface sediments of 50 western Amazonian lakes located across a gradient of observed human disturbance. We will test whether the mismatches of CCMs can predict observed human influence, and whether multi-proxy approaches including charcoal and phytoliths (an additional vegetation proxy) improve model performance, using regression-based approaches. We will apply the best-fit CCM to three paleoecological sequences spanning the last 2000 years at 20-40 year intervals to quantify the role of humans in shaping forest composition in the western Amazonian biodiversity hotspot. Our study will provide novel methodologies and results with relevance across scientific disciplines, and to conservation, policy-making, and Amazonian cultural heritage.


Project number


Main applicant

Dr. C.N.H. McMichael

Affiliated with

Universiteit van Amsterdam, Faculteit der Natuurwetenschappen, Wiskunde en Informatica, Instituut voor Biodiversiteit en Ecosysteem Dynamica - IBED


01/01/2018 to 31/12/2020