Abstract/Summary

Despite nearly a century of debate, scholars continue to disagree about how extensively pre-Columbian (pre- 1492 CE) societies modified and ‘domesticated’ the landscapes of Amazonia. Some view Amazonia as one of the Earth’s last true wildernesses, whereas others argue that pre-Columbian societies have altered its landscapes so extensively that little remains untouched today. Some of the strongest evidence for Amazonian landscape domestication comes from the seasonally flooded savannas of lowland Bolivia, where the pre-Columbian ‘Casarabe Culture’ constructed hundreds of habitation mounds, interconnected by a dense causeway-canal network. Research indicates this culture sustained itself by practicing intensive maize-based agriculture, but little is known about how they utilised the surrounding forest and savanna ecosystems. Employing a dual data-modelling approach that combines agent-based modelling with traditional palaeoecological techniques, this thesis aims to constrain the impact of the Casarabe Culture’s activities on these ecosystems. Agent-based modelling outputs suggest that, while the Casarabe Culture was capable of developing into a large, sedentary society with an estimated population of 10,000 – 100,000, their environmental impact appears to have been relatively localised. This restricted impact stemmed not from an inability to convert land, but rather from the relative availability of farmable land, and their preference for intensively cultivating the most desirable areas. The results of palaeoecological analysis at Laguna Loma Suarez, a small oxbow lake adjacent to one of the Casarabe Culture’s mounds, provides additional evidence of local-scale human land use. This record reveals that while some subsistence activities were practiced in a variety of environmental settings (e.g., maize-based agriculture), others varied spatially (e.g., the utilisation of fire). These results highlight the vital role played by human agency in shaping the human-environment interactions taking place on a landscape. It is not sufficient to merely determine if humans could modify their surroundings; understanding their underlying motivations is also essential.