Domestication of Amazonian forests

Carolina Levis

Research output: Thesisinternal PhD, WU


The idea that Amazonian forests have been largely untouched by humans has fascinated naturalists, policy makers, the media, and natural and social scientists worldwide. For many decades, ecological studies overlooked the influence of past peoples in modern forests. However, humans arrived in the Amazon basin at least 13,000 years Before Present (BP) and populations expanded strongly around 2,500 years BP. Evidence of past human activities has been found in extensive areas previously considered pristine. Anthropogenic soils (Amazonian Dark Earths - ADE) and human-made earthworks found across the basin are examples of the landscapes domesticated by pre-Columbian peoples and evidence of large societies with considerable capacity for modifying the environment. Cultivation and management of Amazonian fora by past societies may have significantly contributed to the ecological patterns we see today.

These findings stimulated an academic debate about the pristineness versus domestication of Amazonian forests. Although most scientists agree that human impacts were spatially heterogeneous across the basin, the scale of past human influences in Amazonian forests remains controversial. A more realistic and broad- scale view is required (Chapter 1). In my thesis, I integrated data from different disciplines in the social and natural sciences to generate the first broad-scale assessment of the effects of long-term human influences in modern Amazonian forests and how these legacies are maintained by local management practices. I studied the Amazon forest as a mosaic of patches domesticated to different degrees by human-nature interactions.

To estimate the influence of past people on modern forests, I first compared the density of sedentary pre-Columbian occupation sites with ADE along some stretches of major white-water rivers and their black or clear-water tributaries that have been sampled by archaeologists. I counted the number of archaeological sites along 12-km sections for both river orders (major and tributary) and found the same density of sites along both orders, showing that archaeological sites are widespread across tributary rivers of the Central Amazon basin (Chapter 2). This result suggests that the influence of past societies in Amazonian landscapes is more extensive than previously imagined and deserves further investigation.

To unravel the effects of long-term human actions at the basin-wide scale, we investigated the relationship between the richness and abundance of 85 domesticated plant species found in Amazonian forests and the distribution of known archaeological sites (Chapter 3). We focused on domesticated species because they are known to have been propagated and selected by peoples in Amazonia and elsewhere in the Americas for food or other uses for a long time. I correlated data from more than 1,000 floristic inventories of the Amazon Tree Diversity Network (ATDN) with a map of more than 3,000 archaeological sites across different Amazonian geological regions compiled by the AmazonArch Network. Our analysis also incorporated environmental data to distinguish the relative importance of environmental conditions from past human factors on modern plant communities. We found that domesticated species were five times more likely to be common in floristic inventories than non- domesticated species and sometimes more abundant far from the places where they were domesticated, suggesting past human dispersal. The richness and abundance of these domesticated species increase with the proximity to archaeological sites and in areas with poorly drained soils and higher rainfall seasonality. Our results show that plant communities in Amazonia are structured by both natural and cultural processes, and refute the idea that these forests are largely untouched by humans.

To understand the relative contribution of past and recent human activities in shaping these current floristic patterns, we expanded our previous analyses to incorporate the influence of current activities (Chapter 4). We found that old-growth forests were transformed by both past and current peoples, but we showed that the effects of recent activities have a smaller role when compared to the persistent effects of pre-Columbian activities on forest composition. Overall, these new analyses strengthened the importance of ancient peoples in explaining the richness and abundance of domesticated species across Amazonia.

In Chapter 5, we investigated how Amazonian people enriched plant communities with useful and domesticated species. To answer this question, we collected extensive information from the literature and data in the field about how Amazonian peoples manage forest resources. With this information, we developed a conceptual model that showed eight key categories of forest management practices that alter natural ecological processes and transform pristine into domesticated forests. Our model allows inferences about how human societies developed ways to interfere with natural ecological processes through time, which created more productive and useful forests across the basin. This long-term process resulted in numerous and diverse patches of useful trees and palms around archaeological sites where humans have lived for centuries or millennia. Thus, a diverse assemblage of useful plant species persists in Amazonian forests due to long-term management practices (Chapters 3-5).

In Chapter 6, we compared the effect of ancient and recent management practices on Amazonian forest soils and vegetation at different distances from pre-Columbian and contemporary villages settled in protected areas. We found that soil nutrients of old-growth forests increased with the proximity to ancient villages, but did not increase with the intensity of recent management activities. By enriching soil nutrients in and around their villages, past societies provided the conditions for forests enriched with species of great interest to modern Amazonian societies. Overall, our results support the hypothesis that ancient management practices have a stronger influence in soils of old-growth forests than recent management practices.

This thesis reveals the persistence of a cultural heritage in modern Amazonian forests, which was created by ancient societies and maintained by present-day peoples. During the millennia that humans have lived in Amazonia, they interacted with nature, modifying landscapes around their villages into forest mosaics formed by patches rich in fertile anthropogenic soils and forest resources, such as foods, medicines and construction materials. To conclude, Amazonian forests hold legacies of past human activities that can only be fully understood by interdisciplinary studies and that require local management practices to be maintained through time.

Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • Wageningen University
  • Bongers, Frans, Promotor
  • Pena Claros, Marielos, Co-promotor
  • Costa, F.R.C., Co-promotor, External person
  • Clement, C.R., Co-promotor, External person
Award date14 Jun 2018
Place of PublicationWageningen
Print ISBNs9789463438735
Publication statusPublished - 2018


  • Cum laude

Fingerprint Dive into the research topics of 'Domestication of Amazonian forests'. Together they form a unique fingerprint.

  • Projects

    Cite this