Fábio Zuker, associate researcher at the Pensi Institute (Fundação José Luiz Setúbal), and the University of São Paulo Institute of Brazilian Studies (IEB)
Bananal Island lies at the meeting point of the Amazon and the savannah-Cerrado. Surrounded by forests, rivers, wetlands, and floodplains, it is a landscape that evokes abundance. Yet this landscape conceals a profound transformation in the region’s hydrology. For many Indigenous communities living on and around the island, between two of the most biodiverse biomes on earth, daily life is increasingly marked by drought, food insecurity, and mosquito-borne diseases.
Location of Bananal Island in Brazil and South America. The dot indicates Bananal Island, and the blue area represents the state of Tocantins, where the island is located. Map created by the author using Datawrapper.
Over the past decades, large-scale irrigation schemes, canals, dykes, and river diversions built to support industrial agriculture have reorganised the circulation of water across the landscape. In doing so, they have reshaped local food systems and disease ecologies: fish populations have declined, floodplain fertility has weakened, locally produced food has increasingly been replaced by increasing dependence on supermarket food from nearby towns, and stagnant water has created favourable conditions for mosquito-borne diseases more commonly associated with precarious urban neighbourhoods across Latin America than with forested Indigenous territories.
Dykes with pumps in the region of the Bananal Island. Photo by Avener Prado
Urbanisation is often understood as the growth of cities, characterised by expanding urban settlements, increasing population density, rural-to-urban migration, and the concentration of industrial and service activities. From this perspective, urbanisation is primarily a demographic, spatial, and economic process, measured through the physical expansion of cities and the transformation of surrounding rural landscapes.
Yet Bananal Island suggests a different way of thinking about urbanisation. Or, more specifically, it suggests that urbanisation can transform territories without making them urban in any conventional sense. These territories become neither forest villages nor cities, but hydrological borderlands of infrastructural urbanisation: spaces where agribusiness infrastructure generates overlapping forms of vulnerability without extending urban citizenship.
Houses in the Takaywrá village of the Krahô people, Lagoa da Confusão, Tocantins. The white tarps covering the house were made from seed bags donated or exchanged with workers from nearby farms. Photo: Avener Prado.
This unprecedented transformation of the region started with the Rio Formoso Project, established in 1979 during Brazil’s military dictatorship as one of the largest irrigated agricultural schemes around Bananal Island. The project helped consolidate intensive rice, soy, corn, and watermelon production through an extensive system of canals, pumps, dams, drainage works, and river diversions. Its expansion was inseparable from the process of concentrating Indigenous peoples into restricted territories and the conversion of surrounding lands into an agribusiness frontier.
This means that while these irrigation projects were designed to transform the region into a hub of intensive agriculture, for Indigenous communities they have also brought deforestation and repeated displacement. As grain farms expanded into forested areas, dams and canals also reshaped the river systems on which local livelihoods depend. These landscape transformations interact with human-induced climate change, as rising temperatures and increasingly prolonged periods of low rainfall further reduce local communities’ capacity to produce food.
Beyond reducing water availability, these transformations have also degraded water quality. According to the Permanent Peoples’ Tribunal in Defence of the Cerrado Territories, altered river dynamics have undermined floodplain agriculture, while fish populations—the primary source of protein for local communities—have declined sharply and face increasing pesticide contamination.
These dynamics become visible across three Indigenous communities on and around Bananal Island. In Boto Velho, a Javaé village on the island, declining river levels and fish stocks have increased dependence on food purchased in the nearby town of Lagoa da Confusão, affecting even ritual exchanges traditionally sustained by local production. In Takaywrá, a Krahô-Takaywrá village, repeated displacement, restricted access to territory, and changes in the Formoso River have undermined fishing and cultivation, while seasonal flooding continues to force families to relocate. In Catàmjê, a Krahô-Kanela village established in 2021 amid soy fields and irrigation canals, flood-control infrastructure has generated new health risks: a ditch built to contain water became a stagnant reservoir during the dry season, contributing to mosquito proliferation and increasing exposure to diseases such as dengue, Zika, chikungunya, and malaria.
A distinct form of urbanization
The ecological transformations and health challenges that the inhabitants of Bananal Island and its surrounding areas face invite us to rethink what counts as urban. Here, a distinct form of urbanisation unfolds, one in which Indigenous territories do not become cities in any conventional sense yet are increasingly shaped by urban-industrial processes. These become spaces in between, neither fully forest nor urban.
Rather than producing urban expansion, these processes generate urban forms of vulnerability in forested communities. Growing dependence on industrialised food systems, declining fisheries, stagnant water, and heightened exposure to mosquito-borne diseases increasingly resemble conditions associated with precarious urban environments. These Indigenous territories are incorporated into urban-industrial metabolism without becoming beneficiaries of the forms of urban citizenship usually associated with improved access to healthcare, education, and other public services. This dynamic resonates with Elizabeth Povinelli’s account of the unequal distribution of harms and benefits of extractivism (Povinelli 2021): the benefits generated by irrigated agriculture circulate to nearby cities and distant commodity markets in the form of cheap protein, while ecological degradation and health risks remain locally concentrated.
The Bananal Island case also contributes to broader debates on planetary urbanisation (Brenner 2018), and the changing boundaries of the urban (Brenner and Schmid 2015). Rather than asking whether there is still an “outside” to urbanisation, it invites us to examine how urbanising processes produce spaces that absorb the ecological and epidemiological costs of urban-industrial development while becoming increasingly shaped by urban dynamics without becoming cities themselves. Bananal Island therefore invites us to reconsider what counts as urban (Castriota and Tonucci 2018). It suggests that urbanisation should be understood not only as the expansion of cities, but also as a process of infrastructural urbanisation capable of transforming river flows, food systems, and disease ecologies far beyond urban boundaries.
As climate change, hydraulic infrastructure, and agricultural expansion continue to reshape landscapes across South America, understanding these forms of infrastructural urbanisation will become increasingly important for recognising where urban vulnerability is being produced—and for whom.
* This work began as a journalistic investigation by the author. Avener Prado contributed the photographs and participated in the first reporting trip, which later developed into a bilingual public-facing piece in Portuguese for O Joio e o Trigo and in English for The Guardian, before being further developed into the academic argument presented here.
Below is the link to the hybrid seminar presented by Fabio on 12th June 2026- “Urbanisation Without Cities? Hydraulic Infrastructure and the Ecologies of Vulnerability in the Amazon”.