Modeling arsenic content in Brazilian soils: What is relevant?

Arsenic accumulation in the environment poses ecological and human health risks. A greater knowledge about soil total As content variability and its main drivers is strategic for maintaining soil security, helping public policies and environmental surveys. Considering the poor history of As studies in Brazil at the country's geographical scale, this work aimed to generate predictive models of topsoil As content using machine learning (ML) algorithms based on several environmental covariables representing soil forming factors, ranking their importance as explanatory covariables and for feeding group analysis. An unprecedented databank based on laboratory analyses (including rare earth elements), proximal and remote sensing, geographical information system operations, and pedological information were surveyed. The median soil As content ranged from 0.14 to 41.1 mg kg−1 in reference soils, and 0.28 to 58.3 mg kg−1 in agricultural soils. Recursive Feature Elimination Random Forest outperformed other ML algorithms, ranking as most important environmental covariables: temperature, soil organic carbon (SOC), clay, sand, and TiO2. Four natural groups were statistically suggested (As content ± standard error in mg kg−1): G1) with coarser texture, lower SOC, higher temperatures, and the lowest TiO2 contents, has the lowest As content (2.24 ± 0.50), accomplishing different environmental conditions; G2) organic soils located in floodplains, medium TiO2 and temperature, whose As content (3.78 ± 2.05) is slightly higher than G1, but lower than G3 and G4; G3) medium contents of As (7.14 ± 1.30), texture, SOC, TiO2, and temperature, representing the largest number of points widespread throughout Brazil; G4) the largest contents of As (11.97 ± 1.62), SOC, and TiO2, and the lowest sand content, with points located mainly across Southeastern Brazil with milder temperature. In the absence of soil As content, a common scenario in Brazil and in many Latin American countries, such natural groups could work as environmental indicators. [Display omitted] •Arsenic concentrations were quantified in topsoil from Brazilian soils.•As was accurately predicted by Random Forest and environmental covariables.•Agricultural soils do not promoted relevant As enrichment as expected.•Temperature, organic carbon, clay, sand, and TiO2 were variables ranked in As modeling.•Four natural groups were suggested, working as environmental indicators of As content.