Reductive solubilization of arsenic in a mining-impacted river floodplain: Influence of soil properties and temperature.

Mining activities have contaminated many riverine floodplains with arsenic (As). When floodplain soils become anoxic under water-saturated conditions, As can be released from the solid phase. Several microbially-driven As solubilization processes and numerous influential factors were recognized in the past. However, the interplay and relative importance of soil properties and the influence of environmental factors such as temperature remain poorly understood, especially considering the (co)variation of soil properties in a floodplain. We conducted anoxic microcosm experiments at 10, 17.5, and 25 °C using 65 representative soils from the mining-impacted Ogosta River floodplain in Bulgaria. To investigate the processes of As solubilization and its quantitative variation we followed the As and Fe redox dynamics in the solid and the dissolved phase and monitored a range of other solution parameters including pH, Eh, dissolved organic C, and dissolved Mn. We related soil properties to dissolved As observed after 20 days of microcosm incubation to identify key soil properties for As solubilization. Our results evidenced reductive dissolution of As-bearing Fe(III)-oxyhydroxides as the main cause for high solubilization. The availability of nutrients, most likely organic C as the source of energy for microorganisms, was found to limit this process. Following the vertical nutrient gradient common in vegetated soil, we observed several hundred µM dissolved As after 1-2 weeks for some topsoils (0-20 cm), while for subsoils (20-40 cm) with comparable total As levels only minor solubilization was observed. While high Mn contents were found to inhibit As solubilization, the opposite applied for higher temperature (Q10 2.3-6.1 for range 10-25 °C). Our results suggest that flooding of nutrient-rich surface layers might be more problematic than water-saturation of nutrient-poor subsoil layers, especially in summer floodings when soil temperature is higher than in winter or spring.

Soil-to-plant transfer of arsenic and phosphorus along a contamination gradient in the mining-impacted Ogosta River floodplain.

Riverine floodplains downstream of active or former metal sulfide mines are in many cases contaminated with trace metals and metalloids, including arsenic (As). Since decontamination of such floodplains on a large scale is unfeasible, management of contaminated land must focus on providing land use guidelines or even restrictions. This should be based on knowledge about how contaminants enter the food chain. For As, uptake by plants may be an important pathway, but the As soil-to-plant transfer under field conditions is poorly understood. Here, we investigated the soil-to-shoot transfer of As and phosphorus (P) in wild populations of herbaceous species growing along an As contamination gradient across an extensive pasture in the mining-impacted Ogosta River floodplain. The As concentrations in the shoots of Trifolium repens and Holcus lanatus reflected the soil contamination gradient. However, the soil-to-shoot transfer factors (TF) were fairly low, with values mostly below 0.07 (TF=Asshoot/Assoil). We found no evidence for interference of As with P uptake by plants, despite extremely high molar As:P ratios (up to 2.6) in Olsen soil extracts of the most contaminated topsoils (0-20cm). Considering the restricted soil-to-shoot transfer, we estimated that for grazing livestock As intake via soil ingestion is likely more important than intake via pasture herbage.

Bioaccessibility of arsenic in mining-impacted circumneutral river floodplain soils.

Floodplain soils are frequently contaminated with metal(loid)s due to present or historic mining, but data on the bioaccessibility (BA) of contaminants in these periodically flooded soils are scarce. Therefore, we studied the speciation of As and Fe in eight As-contaminated circumneutral floodplain soils (≤ 21600 mg As/kg) and their size fractions using X-ray absorption spectroscopy (XAS) and examined the BA of As in the solids by in-vitro gastrointestinal (IVG) extractions. Arsenopyrite and As(V)-adsorbed ferrihydrite were identified by XAS as the predominant As species. The latter was the major source for bioaccessible As, which accounted for 5-35% of the total As. The amount of bioaccessible As increased with decreasing particle size and was controlled by the slow dissolution kinetics of ferrihydrite in the gastric environment (pH 1.8). The relative BA of As (% of total) decreased with decreasing particle size only in a highly As-contaminated soil--which supported by Fe XAS--suggests the formation of As-rich hydrous ferric oxides in the gastric extracts. Multiple linear regression analyses identified Al, total As, C(org), and P as main predictors for the absolute BA of As (adjusted R(2) ≤ 0.977). Health risk assessments for residential adults showed that (i) nearly half of the bulk soils may cause adverse health effects and (ii) particles <5 µm pose the highest absolute health threat upon incidental soil ingestion. Owing to their low abundance, however, health risks were primarily associated with particles in the 5-50 and 100-200 µm size ranges. These particles are easily mobilized from riverbanks during flooding events and dispersed within the floodplain or transported downstream.

Arsenic species formed from arsenopyrite weathering along a contamination gradient in Circumneutral river floodplain soils.

Arsenic is a toxic trace element, which commonly occurs as contaminant in riverine floodplains and associated wetlands affected by mining and ore processing. In this study, we investigated the solid-phase speciation of As in river floodplain soils characterized by circumneutral pH (5.7-7.1) and As concentrations of up to 40.3 g/kg caused by former mining of arsenopyrite-rich ores. Soil samples collected in the floodplain of Ogosta River (Bulgaria) were size-fractionated and subsequently analyzed using a combination of X-ray fluorescence (XRF) spectrometry, powder X-ray diffraction (XRD), X-ray absorption spectroscopy (XAS), and selective chemical extraction of poorly crystalline mineral phases. Arsenic and Fe were found to be spatially correlated and both elements were strongly enriched in the fine soil particle size fractions (<2 µm and 2-50 µm). Between 14 and 82% of the total As was citrate-ascorbate extractable. Molar As/Fe ratios were as high as 0.34 in the bulk soil extracts and increased up to 0.48 in extracts of the fine particle size fractions. Arsenic K-edge XAS spectra showed the predominance of As(V) and were well fitted with a reference spectrum of As(V) adsorbed to ferrihydrite. Whereas no As(III) was detected, considerable amounts of As(-I) were present and identified as arsenopyrite originating from the mining waste. Iron K-edge XAS revealed that in addition to As(V) adsorbed to ferrihydrite, X-ray amorphous As(V)-rich hydrous ferric oxides ("As-HFO") with a reduced number of corner-sharing FeO6 octahedra relative to ferrihydrite were the dominating secondary As species in the soils. The extremely high concentrations of As in the fine particle size fractions (up to 214 g/kg) and its association with poorly crystalline Fe(III) oxyhydroxides and As-HFO phases suggest a high As mobilization potential under both oxic and anoxic conditions, as well as a high bioaccessibility of As upon ingestion, dermal contact, or inhalation by humans or animals.

Quantifying sediment-associated metal dispersal using Pb isotopes: application of binary and multivariate mixing models at the catchment-scale.

In this study Pb isotope signatures were used to identify the provenance of contaminant metals and establish patterns of downstream sediment dispersal within the River Maritsa catchment, which is impacted by the mining of polymetallic ores. A two-fold modelling approach was undertaken to quantify sediment-associated metal delivery to the Maritsa catchment; employing binary mixing models in tributary systems and a composite fingerprinting and mixing model approach in the wider Maritsa catchment. Composite fingerprints were determined using Pb isotopic and multi-element geochemical data to characterize sediments delivered from tributary catchments. Application of a mixing model allowed a quantification of the percentage contribution of tributary catchments to the sediment load of the River Maritsa. Sediment delivery from tributaries directly affected by mining activity contributes 42-63% to the sediment load of the River Maritsa, with best-fit regression relationships indicating that sediments originating from mining-affected tributaries are being dispersed over 200 km downstream.