Artisanal and small-scale mining (ASM) in the north region of Brazil, especially alluvial gold panning, has declined since 1985 but is still responsible for pollution by toxic metals in and near the Madeira River, which is the largest tributary of the Amazon River.
A study conducted by researchers at São Paulo State University (UNESP) in Rio Claro, Brazil, in partnership with colleagues at Queensland University of Technology in Australia found relatively high levels of mercury from gold panning that were accumulated in sediment in lakes in the Madeira Basin.
The results of the study, which was supported by FAPESP under the aegis of its São Paulo Researchers in the International Collaboration Program (SPRINT), were published in the journal Ecotoxicology and Environmental Safety. Researchers affiliated with the Federal University of Rondônia (UNIR) in Brazil and Shenzhen University in China also participated in the study.
“Although the intensity of ASM and gold panning on the Madeira River has diminished over the last two decades, this activity remains the leading source of the mercury we found in lake sediments in the basin in question,” Daniel Marcos Bonotto, Full Professor at UNESP Rio Claro and first author of the study, told Agência FAPESP.
The project was the second for which Bonotto was the principal investigator with support from FAPESP via SPRINT. The first project was in 2016, when he collaborated with Trevor Elliot, a faculty member at Queen’s University Belfast (UK), for research on environmental tracers for water resource management.
“SPRINT favors mobility and identification of projects in collaboration with researchers in other countries, even if the projects haven’t yet been formatted,” Bonotto said.
The creation of new research partnerships is one of the precise aims of SPRINT, which has been running for five years. Launched in April 2014, the program offers seed funding for medium- to long-term cross-border collaborative projects involving researchers affiliated with universities and research institutions in São Paulo State.
“FAPESP’s expectation is that the provided seed funding, plus a matching contribution from the partner university, will enable researchers to interact in a project while simultaneously developing a collaboration that leads to a medium- or long-term proposal for a joint research project to be submitted to FAPESP and made accessible to the overseas agencies and researchers involved,” said Carlos Henrique de Brito Cruz, FAPESP’s Scientific Director.
According to Marilda Solon Teixeira Bottesi, FAPESP’s special advisor for research collaboration, SPRINT aims to consolidate existing research partnerships or stimulate new collaborations by funding scientific missions.
“Through the projects supported by SPRINT, participating researchers have the opportunity to visit institutions and familiarize themselves with the laboratories and research projects led by their partners. This knowledge then serves as a basis for proposing joint projects,” she said.
Bonotto’s partnership with Australian researchers, for example, made it possible to expand the investigation begun by his group at UNESP in the 1990s, when it started to collect samples from lake sediments at different depths and rock formations surrounding the lakes along the Madeira River in the vicinity of Porto Velho, Rondônia State's capital city, to measure the mercury levels and determine where the toxic metal originated from.
Mercury is a human health hazard when ingested via fish consumption. Mercury naturally contaminates rivers via leaching from the soil or atmospheric emissions from Andean volcanoes. Rivers are anthropogenically contaminated by mercury via gold panning.
“During our fieldwork, we actually witnessed artisanal gold miners dumping mercury into lakes on the Madeira River,” Bonotto said.
To estimate the contributions of natural sources and alluvial mining to the levels of mercury found in the region, the researchers analyzed rock and sediment samples from nine lakes in the Madeira Basin using Bayesian networks.
Bayesian networks are a type of probabilistic graphical model that represents the relations of causality among variables in a system. These models are increasingly used to understand complex environmental systems. Examples include the prediction of species abundance as a function of habitat characteristics.
“Our colleagues in Australia are experts in Bayesian network modeling and found it interesting to use this approach to try to assess the contributions of the different sources of mercury that pollute these lakes in the Madeira Basin based on the data from the sediments we collected,” Bonotto said.
Their analysis showed that while a large amount of this mercury comes from geological formations and soil in Amazon ecosystems, alluvial gold panning also accounts for a substantial proportion. The researchers found significantly higher levels of mercury in lake sediment than in rock samples, implying that most of the mercury pollution in the lakes comes from gold panning.
Given the decline in gold panning on the Madeira River in recent years, the high levels of sediment contamination are likely due to past emissions of mercury, according to the authors of the study.
“Lake bottom sediments typically retain a great deal of evidence of pollution,” Bonotto said. “The sediment columns we collected at different depths, for example, recording the mercury pollution over many years.”
Bonotto now plans to use the same statistical approach in collaboration with the group at Queensland University of Technology in Australia to study the environmental impact of the footwear industry in Franca, São Paulo State. This is an important industry in Brazil and is a source of chromium pollution in local rivers.
“We also want to evaluate the influence of nitrate on the groundwater transportation of uranium to the Guarani aquifer in São Paulo State,” Bonotto said.
Each year, SPRINT holds four calls for proposals, which must be submitted by the last Monday in January, April, July and October. The first call in 2019, for which the submission deadline is April 29, has set a new record for the number of participating institutions. Sixteen institutions in various countries have applied so far.
To date, SPRINT has supported more than 100 projects conducted by researchers affiliated with public and private universities and research institutions in São Paulo State in collaboration with colleagues from Argentina, Australia, Belgium, Canada, Chile, China, Denmark, England, France, Germany, Iran, Japan, the Netherlands, Northern Ireland, Scotland, South Africa, Spain, Sweden, the United States, and Wales.
The SPRINT 1/2019 call for proposals is published at www.fapesp.br/en/12514.
The article “Assessing mercury pollution in Amazon River tributaries using a Bayesian network approach” (DOI: 10.1016/j.ecoenv.2018.09.099) by Daniel Marcos Bonotto, Buddhi Wijesiri, Marcelo Vergotti, Ene Glória da Silveira and Ashantha Goonetilleke can be retrieved by subscribers to the journal Ecotoxicology and Environmental Safety from www.sciencedirect.com/science/article/pii/S0147651318309734.