The distribution of arsenic in soils affected by the Aznalcóllar mine spill, SW Spain

In April 1998, a holding lagoon containing pyrite ore processing waste, failed and released an estimated 5-6 million m(3) of metal rich sludge and acidic waters into the Rio Guadiamar, SW Spain. Over 2700 ha of the internationally important Doñana National and Natural Parks were contaminated. The sludge component of the waste contained up to 0.6% arsenic. This paper presents an extensive set of 0-5 cm soil analyses results from samples taken 6-8 months after the spill. Data are presented on pseudo-total arsenic levels in these samples, and on arsenic removed by both single batch and sequential extraction techniques. Pseudo-total levels of arsenic in the sludge ranged from 1521 to 3510 mg kg(-1), and a mean 4.4% of this was found to be extractable using 2.5% acetic acid. Soils in the Guadiamar Valley and Entremuros areas (those worst affected by the spill) were found to contain 85.4-782 mg kg(-1) and 7.1-196 mg kg(-1) pseudo-total arsenic, respectively. The mean acetic acid extractable component in these areas was 2.5% and 4.9%, respectively. Background pseudo-total arsenic levels in the soils of the area have been found to be 4.2-13.6 mg kg(-1). Rapid input of acidic waters, and the acidic nature of the sludge itself, may have caused depletion of Mg, Na and K, and to a lesser extent Mn, Ca and PO(4) in the upper 5 cm of the worst affected soils. Sequential extraction results suggest clear As-Fe associations, and possible As-Mn associations within the more soluble phases.

Field fluxes and speciation of arsines emanating from soils

The biogeochemical cycle of arsenic (As) has been extensively studied over the past decades because As is an environmentally ubiquitous, nonthreshold carcinogen, which is often elevated in drinking water and food. It has been known for over a century that micro-organisms can volatilize inorganic As salts to arsines (arsine AsH(3), mono-, di-, and trimethylarsines, MeAsH(2), Me(2)AsH, and TMAs, respectively), but this part of the As cycle, with the exception of geothermal environs, has been almost entirely neglected because of a lack of suited field measurement approaches. Here, a validated, robust, and low-level field-deployable method employing arsine chemotrapping was used to quantify and qualify arsines emanating from soil surfaces in the field. Up to 240 mg/ha/y arsines was released from low-level polluted paddy soils (11.3 ± 0.9 mg/kg As), primarily as TMAs, whereas arsine flux below method detection limit was measured from a highly contaminated mine spoil (1359 ± 212 mg/kg As), indicating that soil chemistry is vital in understanding this phenomenon. In microcosm studies, we could show that under reducing conditions, induced by organic matter (OM) amendment, a range of soils varied in their properties, from natural upland peats to highly impacted mine-spoils, could all volatilize arsines. Volatilization rates from 0.5 to 70 µg/kg/y were measured, and AsH(3), MeAsH(2), Me(2)AsH, and TMAs were all identified. Addition of methylated oxidated pentavalent As, namely monomethylarsonic acid (MMAA) and dimethylarsinic acid (DMAA), to soil resulted in elevated yearly rates of volatilization with up to 3.5% of the total As volatilized, suggesting that the initial conversion of inorganic As to MMAA limits the rate of arsine and methylarsines production by soils. The nature of OM amendment altered volatilization quantitatively and qualitatively, and total arsines release from soil showed correlation between the quantity of As and the concentration of dissolved organic carbon (DOC) in the soil porewater. The global flux of arsines emanating from soils was estimated and placed in the context of As atmospheric inputs, with arsines contributing from 0.9 to 2.6% of the global budget.