Charcoal maps ten thousand years of fire in Amazonia

Before satellites, Amazonian fire history accumulated in soil charcoal and burned material from archaeological sites. Crystal N. H. McMichael, Britte M. Heijink, Nina H. Witteveen, Annabel Zwarts and Mark B. Bush assembled that evidence across ten millennia. The paper appeared in Frontiers of Biogeography, published by Pensoft Publishers; Pensoft’s release distributed by EurekAlert carries no named individual reporter byline.

The dataset contains 1,345 calibrated radiocarbon dates from 303 sites. Radiocarbon, or carbon-14, is a radioactive form of carbon whose predictable decay helps estimate the age of organic matter; calibration corrects that estimate for atmospheric variation. Mapping the probabilities produced a pyrogeography, meaning the spatial and temporal distribution of fire.

The sequence does not begin in the forest center. The earliest records cluster along Amazonian margins and the main river channel. Geographic spread becomes visible about 6,000 years ago, and by 2,000 to 3,000 years ago burned material occurs in every analyzed region, including the interior. The authors link the pattern to settlement, cultivation and land management rather than a uniform wave of natural fires.

A two-step decline then changes the timeline. Date frequencies fall roughly 700 to 500 years ago, before European arrival, and again during the last four centuries, when epidemics, violence and displacement associated with colonization reduced Indigenous populations. The first drop is consistent with early abandonment of some places; the second with the demographic collapse often called the Great Dying.

This does not make all Amazonia an artificial landscape or equate ancient Indigenous burning with present-day deforestation. The method measures the relative occurrence of datable material, not burned area, fire intensity or exact population. Regions investigated more heavily can loom larger in the record, and charcoal at one point cannot represent the entire basin.

The long view still changes the conservation question. A forest may carry legacies of human management while remaining vulnerable to modern fires intensified by clearing, fragmentation and warming. Earlier reviews by Mark Bush and colleagues stressed that natural fire is rare in the wet forest core and that climate and land use interact to determine risk.

The evidence supports a conclusion that human-set fire spread unevenly through Amazonia during the Holocene — the geological epoch beginning about 11,700 years ago — and helped structure modern forests, especially over the last two millennia. It does not quantify total transformed forest or justify transplanting old practices into today’s conditions. The key uncertainty is disentangling population, climate and sampling bias region by region.