Ancient human impacts in Amazonia – the debate continues…

How well can you ever really know 5.5 million km2 of hyperdiverse forest? The Amazon region, and this question, are at the heart of several ongoing debates in the natural sciences – why are there so many species? how much carbon can the forest store? how much did ancient humans impact the forests, and can we still see their effects today?

How fully can you really know a forest as huge and diverse as the Amazon?

This last question was the subject of a paper by Caroline Levis and colleagues published in the prestigious journal Science in March this year, which caused significant ripples within and outside the scientific community. In a nutshell, the article found that tree species humans have cared for are five times more likely to be forest ‘hyperdominants’ than you’d expect. Past human impacts also help explain where these species are found now, accounting for up to 20% of the variation in their distribution (for comparison, environmental factors explained up to 30%). The paper’s conclusion, as picked up by the media, is that “modern tree communities in Amazonia are structured to an important extent by a long history of domestication by Amazonian peoples.”

So, debate settled? Not quite.

Theobroma cacao – the tree cocoa beans come from – was one human-influenced species in the Levis et al study

Responding to the paper some critics pointed out that the potential influence of soil nutrients hadn’t been adequately considered; others suggested the data showed that human effects were only significant within 20km of archaeological sites; and yet another argued that most of the human-influenced trees are so short-lived that pre-colonial individuals would be dead by now. Furthermore, another study published around the same time showed these Amazon forest plots are disproportionately close to places highly affected by ancient humans, so maybe it was inevitable that the sampled areas would still be showing their influence.

This, perhaps, goes to the heart of the problem with trying to uncover the human history of all Amazonia: there isn’t much detailed region-wide environmental data, the non-random tree plots cover a vanishingly small fraction of the forest, and there are many archaeological sites we may never find. Reflecting on the paper and its findings at a recent TPRG meeting, it was sobering to consider how this can limit the conclusions of wide-ranging studies like this.

It was also striking how, when it comes to ancient human impacts on forests, people can draw very different conclusions from the same data. As I embark on my research into the effects of past humans and climate change on southern Brazil’s Araucaria forests, these will certainly be important things to bear in mind!



The original paper by Caroline Levis and colleagues can be found here: Some scientists have responded to the paper in its eLetters section (, and Christopher Dick posted his thoughts on his research group website ( The paper on the link between forest plots and areas of ancient human impact was published by Crystal McMichael et al., and can be found here:

Leave a Reply

Your email address will not be published. Required fields are marked *