Optimisation of arsenic-rich MIne wastes phytostabilisation strategies: prediction of impacts on water and pollutants bioavailability to plants linked with MicrObial activities – oMIMo

These objectives are being addressed by the oMIMo consortium (BRGM, ISTO, Li2D-DRF/CEA, and LEB/ADERA) through an interdisciplinary approach combining geochemistry, numerical modeling, plant physiology, microbiology, and omics techniques, along with strong operational knowledge of legacy mine site management. The oMIMo methodology is based on a controlled assisted phytostabilization scenario. oMIMo proposes to integrate into the reactive transport model (RTM) indicator data on active microbial processes related to arsenic (As), iron (Fe), and sulfur (S) metabolisms.

In March 2023, a mesocosm was filled with 1200 kg of residues from the former tin (Sn) mine located in Abbaretz (Abb). This pilot experiment reproduced, at a metric scale, the different compartments of the deposit: phytostabilized surface, underlying unsaturated zone, and deeper saturated zone. In parallel, sub-metric experiments in 20 L pots began in early May 2023 using two types of arsenic-rich residues: those from Abbaretz and those from the former Ag-Pb mine in Pontgibaud (Pontg). The plant species selected for these trials is Festuca rubra, and the mine residues were amended in the plant colonization zone with a mixture of limestone (2%) and compost (3%).

These experiments were conducted over two years, enabling the collection of detailed data on the evolution of geochemistry, microbial processes, and the bioavailability/toxicity of As and metals for plants. The RTM will be based on mass balance and thermodynamic laws. A network of stoichiometric metabolic reactions will represent the suspected redox sequence occurring in the residues, feeding an initial version of the model with state-of-the-art knowledge and early geochemical data from both metric and sub-metric experiments.

This framework will then be enriched to simulate controlled thermokinetic reactions and to couple microbial metabolisms with abiotic geochemical processes. The reaction network will be refined through the identification of active bacteria via 16S rRNA sequencing from sub-metric experiments, followed by metagenomic and metaproteomic data from the metric-scale experiment. For each detected metabolism, the thermodynamic term will be coupled with a Monod equation, where the “biomass” component is linked to a growth function specific to each metabolic group.

Parameters for these functions will first be estimated by determining the abundance (via MPN, Most Probable Number) of bacteria from key metabolic groups. Omics data will provide a more precise picture of active metabolisms and their distribution from the surface to the saturated zone. A third version of the model will incorporate transport parameters. The calculated data will be compared with measured water geochemistry and plant bioindicators (stress levels and pollutant concentrations) for the surface compartment.

In June 2025, the two oMIMo project experiments—mesocosm and pot trials—were completed following the final sampling campaigns. The pot experiment showed optimal growth of Festuca rubra in sand, moderate growth with Pontg. residue, and poor growth with Abb. residue. Dissolved arsenic concentrations were significantly higher in Abb. pore waters (>1 mg/L) compared to Pontg. pore waters (5–12 µg/L). In the mesocosm (Abb.), pore water arsenic concentrations before phytostabilization were around 50 µg/L.

Prior to any amendment, microbial metabolic succession phenomena were observed in the water-saturated layer of the mesocosm: a drop in NO₃⁻ concentration followed by an increase in dissolved Fe and As. The evolution of pore water parameters in the mesocosm provided insights into changing conditions across layers and the downward diffusion of phytostabilization effects. Geochemical parameters changed immediately in the surface zone, were affected after 50 days in the intermediate zone, and after 200 days in the bottom layer. In this deep zone, AsIII was the dominant arsenic species, and its solubilization was linked to FeII release. An acceleration in arsenic concentration increase was observed after phytostabilization.

In the pots, pH increase and organic matter input from amendments mobilized arsenic. These results align with the observed rise in dissolved As in the upper mesocosm layer post-amendment. MPN analyses indicated that Fe-oxidizing and S-oxidizing bacteria were more abundant in Pontg., while SO₄²⁻-reducing bacteria were more prevalent in Abb. Sequencing revealed differences in active microbial communities between planted and unplanted pots, and between Abb. and Pontg. In Abb., more iron-cycle-related genera were detected in planted pots than in unplanted ones, including ferri-reducing (Citrifermentans) and neutrophilic ferro-oxidizing (Sideroxydans) genera.

Phytotoxicity was assessed using Lactuca sativa as an indicator plant, grown on both amended and unamended solids. The most inhibitory solid, in terms of both germination and biomass production, was Abb. Amendment did not improve biomass production on Abb. residue, whereas it enhanced growth on sand and Pontg. residue. Regarding the omega-3 index, toxicity to lettuce was higher with Abb. than Pontg., regardless of amendment. For Festuca rubra, the omega-3 index showed slightly lower toxicity in amended Abb. compared to amended Pontg. Stress levels were moderate to high with both residues.

Modeling tasks began with identifying biogeochemical reactions likely to provide the most energy to microorganisms.

All bio-indicators on the samples are currently being analyzed. These include markers of oxidative stress: lipid peroxidation (MDA and Omega-3 Index), chlorophylls; non-enzymatic antioxidants: anthocyanins, phenolic compounds, flavonoids, proline, soluble sugars; enzymatic antioxidants: superoxide dismutase, catalase, peroxidase, ascorbate peroxidase, glutathione-S-transferase; and nitrate reductase.

16S metabarcoding sequencing of complementary DNA obtained from RNA extracted from the pot experiment residues at the end of the trial is underway to identify active bacteria in the Abb. and Pontg. residues.

DNA and proteins associated with samples from successive core samplings, currently stored at -80°C, are being extracted. Metagenomic and metaproteomic analyses of microbial communities from these samples will be used to calibrate and improve the reactive transport model (RTM). A postdoctoral researcher has been recruited at CEA-LI2D to contribute to the analysis of metagenomic and metaproteomic data and has already taken charge of the methodology.

One of the risks associated with this experiment is the potentially low abundance of proteins and DNA in the residues, particularly in the unamended ones. Two strategies are proposed to overcome this challenge. For proteomics, the SCP (Single Cell Proteomics) approach may be developed to extract more data from small amounts of protein. Another strategy involves extracting proteins and DNA from larger quantities of material.

In terms of data processing, a functional analysis approach is favored over a purely taxonomic description. This will be cross-referenced with the energy-yielding metabolic reactions identified through thermodynamic modeling efforts.

For the “transport” component of the model, tracer tests will be conducted at the end of 2025 in laboratory columns (500 mL) to precisely determine the hydraulic behavior of Abb. residues under both unsaturated and saturated conditions. A postdoctoral researcher dedicated to modeling tasks has been recruited for an 18-month period.

The oMIMo project has not yet been the subject of scientific production after 6 months of completion.

Securing mining residues represents a major environmental challenge. Most metal mines produced waste containing iron (Fe) and sulfur (S), with the toxic element arsenic (As) being present at more than 50% of sites. Phytostabilisation often appears as an appropriate option to minimize the risks linked to the dispersion of particles by erosion, at a moderate cost. However, its integration into rehabilitation projects by site managers must be supported by a quantitative assessment of its effect on the fate of As and metals in the waste masses. oMIMo aims to develop a tool for predicting the mobility of As and metals and their availability to plants in a phytostabilised arsenic mining waste, based on a reactive transport modeling methodology (RTM) integrating microbial parameters. This objective will be addressed by the oMIMo consortium (BRGM, ISTO, Li2D-DRF/CEA and LEB/ADERA) through an interdisciplinary approach combining geochemistry, numerical modeling, plant physiology, microbiology (classical and molecular) and omics approaches coupled with a good knowledge of the former mining sites operational management. The oMIMo methodology is based on a controlled scenario of assisted phytostabilisation, developed up to the metric pilot and for which a first version of RTM has been developed for the residues of an Ag-Pb mine. oMIMo proposes to integrate into the RTM data indicative of active microbial processes related to As, Fe and S metabolisms. Sub-metric experiments will be carried out on 2 types of residue, that (well known) of the Ag-Pb mine and those, also arsenic, of a tin (Sn) mine. These tests, coupled with a metric pilot carried out with the residue of the Sn mine, will make it possible to acquire information on the evolution of the geochemistry, the microbial processes, and the bioavailability/toxicity of As and metals for plants. The RTM will be based on a mass balance and thermodynamic laws. A network of stoichiometric metabolic reactions will represent the redox sequence suspected to occur in the tailings, feeding a first version of the model with state-of-the-art and geochemical data. In the second version of the model, this formalism will be enriched in order to simulate controlled thermokinetic reactions and to couple microbial metabolisms with abiotic geochemical processes. The reaction network will be corrected via the identification of active bacteria by 16SrRNA sequencing, then by omics data (metagenomics and metaproteomics) from the metric experiment. For each detected metabolism, the thermodynamic term will be coupled with a Monod equation whose “biomass” component will be linked to a growth function for each metabolic group. The parameters of these functions will first be estimated using the “Most Probable Number” of bacteria from metabolic groups. Omics data will provide a more accurate picture of active metabolisms and their distribution from the surface to the water-saturated zone. A third version of the model will integrate transport parameters. The calculated data will be compared with measured water geochemistry data and plant bio-indicators (stress level and pollutant concentrations) for the surface compartment. In terms of impacts, oMIMo will address the need to predict impacts, including arsenic and metal fluxes toward groundwater, when developing phytostabilisation as an operational option for As-rich mining sites management. oMIMo will provide a modeling methodology accompanied by recommendations, including the type of relevant analyses (geochemical, mineralogical and biological) to be carried out for the numerical determination of the medium and long-term evolution of As and metal fluxes.