Legacy mercury in soils from artisanal and small-scale gold mining (ASGM) landscapes in Jardim do Ouro, Pará (Brazilian Amazon): persistence, risk, and retention controls

Mercury (Hg) contamination associated with artisanal and small-scale gold mining (ASGM) is a persistent environmental issue in the Brazilian Amazon. While ASGM is widely recognized as a major source of Hg, elevated concentrations in soils and sediments cannot always be attributed to mining alone. Geogenic enrichment, long-term atmospheric deposition, and landscape processes such as erosion and sediment transport can also contribute substantially to Hg distribution. In highly disturbed Amazonian landscapes, this overlap complicates contamination attribution and can bias risk assessments when regional geochemical baselines are not defined. In this context, this thesis investigates Hg contamination in superficial soils from Jardim do Ouro village (Pará State), comparing three ASGM sites with contrasting activity histories (active, deactivated in January 2024, and deactivated for ~30 years) against a control area without ASGM records used as a regional geogenic baseline. The study integrates total mercury (THg) concentrations and ecological risk indices with particle-size distribution, mineralogical and surface characterization (XRD and ATR-FTIR), thermogravimetric data (TG/DSC), and basic physicochemical properties (WHC and pH). This approach was designed to clarify both the persistence of legacy Hg in post-ASGM landscapes and the soil fractions controlling Hg retention under different disturbance histories. Soils from the ASGM site deactivated for ~30 years still exhibited elevated THg concentrations (27 to 8,862 ng g⁻¹; median = 324 ng g⁻¹) and ecological risk levels above regulatory thresholds, demonstrating that Hg contamination can remain environmentally relevant decades after mining activity has ceased. In contrast, the active and recently deactivated sites generally displayed THg concentrations close to the regional baseline (178 ng g⁻¹), with exceedances largely restricted to localized hotspots directly associated with ASGM operations. These results indicate that current mining activity does not necessarily translate into higher median THg at the site scale, whereas long-term deactivated areas can retain a strong contamination legacy. In the long-term deactivated site, fine particle content explained 73% of the variability in log10-transformed THg, revealing a robust texture control on Hg retention. Mineralogical analyses indicated broadly similar frameworks dominated by quartz and kaolinite-group minerals, consistent with highly weathered tropical soils developed on granitic parent material. However, accessory phases such as gibbsite, mica-type phyllosilicates, smectites, and sporadically occurring feldspars highlighted meaningful sample-scale variability linked to weathering intensity, soil development, and disturbance history. The association between Fe–Mn (oxyhydr)oxides and THg was not uniform across sites, supporting a context-dependent role of reactive phases in Hg retention and redistribution. Thermogravimetric profiles were dominated by mass loss in the 380–600 °C interval, and the strongest between-site contrasts were expressed in this range, indicating that site-related differences are primarily reflected in major soil fractions controlling thermal behavior. THg was systematically associated with thermogravimetric fractions, particularly in the ~30 years deactivated site, where strong correlations across all temperature intervals linked Hg enrichment to multiple soil components. WHC was positively associated with THg only in the long-term deactivated site, consistent with preferential Hg enrichment in fine-textured soils with higher water retention capacity. Soil pH showed no significant relationship with THg in any site. Overall, this thesis demonstrates that Hg risk in ASGM landscapes is not defined solely by the presence of active mining, but by the interaction between contamination legacy, soil texture, physical properties, reactive mineral phases, and disturbance history. These findings reinforce the need for regional geochemical baselines, long-term monitoring, and soil-informed approaches to Hg risk assessment and management in the Brazilian Amazon.