From unseen to seen in post-mining polluted territories: (in)visibilisation processes at work in soil contamination management

The study is situated within debates on a prospective mining revival in metropolitan France, where ecological transition and energy independence discourse legitimize reopening mines. Former mining regions such as the Cévennes face persistent economic, environmental, sanitary, and geotechnical issues, notably polymetallic soil contamination that can be invisible unless made visible through events or measurements. In Saint-Laurent-Le Minier—home to major historic lead–zinc deposits—widespread soil contamination persists, with lead measured up to 30 g/kg (about 600 times typical background). Management actions have included soil replacement and, later, state-led rehabilitation. The article asks how visibilisation and invisibilisation processes operate in research and management of soil contamination, and how lay and expert knowledge interact to shape problem framing, stakeholder inclusion, and technical choices (notably phytoremediation). It emphasizes researchers’ social and political responsibility in co-producing knowledge and influencing remediation trajectories.

The paper mobilizes scholarship on visibility/invisibility in (post-)mining territories (Fontaine, 2016) and risk communication around contamination (Becerra et al., 2016; Doumas et al., 2018), noting that metals and extraction are often rendered invisible or 'dis-origined' (McKay, 2021). It distinguishes material invisibility (microscopic pollutants made visible via instruments) and discursive invisibility (narratives that magnify or minimize risks, Beck, 1992). Rhetorical strategies, such as technological ambiguity in 'green' solutions (Schneider et al., 2016; Paliewicz, 2024), also shape public meaning. Building on Lemaire-Crespy (2013), the authors review how different local actors alternately render pollution visible or invisible: long-term residents pragmatically 'live with' and downplay pollution; newcomers and media heighten visibility; institutions seek image renewal via 'green' identity. STS literature (Latour & Woolgar, 1986; Harding, 2015; Frickel & Vincent, 2007) underscores that limited procedures and dominant actors’ framing generate partial knowledge; co-production perspectives (Jasanoff, 2004) are central to the study’s lens.

Design: Qualitative case study (Jan 2022–Jun 2023) combining documentary analysis, semi-structured interviews, and field observations, with reflexive attention to visibilisation/invisibilisation in the research process itself. - Literature and media review: Identified 17 scientific papers (Web of Science) and 52 media articles (Europresse) covering Saint-Laurent-Le Minier from mine closure to May 2022. Analyzed temporal trends and thematic emphases (e.g., remediation focus; differential media attention). - Interviews (2022–2023): Semi-directive interviews targeted stakeholders familiar with the territory (purposeful sampling per Paille & Mucchieli, 2021). Panel included researchers (microbiology, environmental chemistry) involved in phytoremediation, local inhabitants (including former mayor, former miners, heritage association members), and institutional stakeholders (ARS, DREAL, Géodéris, ADEME). Counts reported: Researchers ≈5, Inhabitants ≈10, Institutional stakeholders ≈6. - Supplementary data: Considered additional phytoextraction projects via secondary sources and testimonies from non-involved researchers and local stakeholders. - Fieldwork: Four field trips with scientists, inhabitants, and students for situated observations of research sites and interactions. - Reflexive stance: Authors note sample visibility bias (greater access to visible projects and engaged inhabitants), limited pool of informed participants (rapid data saturation), hesitancy to speak on record due to past tensions (importance of informal interactions), and possible invisibilisation induced by research choices (questions prioritized/omitted).

- Publication dynamics: Since ~2010, scientific outputs on the site accelerated, specializing in soil contamination and phytoremediation; media coverage did not peak with ADEME’s site assumption and used euphemisms for pollution. Media preferentially spotlighted phytoextraction, largely ignoring phytostabilisation despite its dominance in research and implementation. - Research focus shaping solutions: Early work on metal-resistant plants (e.g., Noccaea caerulescens; Anthyllis vulneraria–Rhizobium metallidurans) at Les Avinières (among the highest Zn levels in Europe) framed soil contamination as a phytoremediation problem, ultimately institutionalizing phytostabilisation. Health impacts were comparatively less visible in academic work (though monitored by state agencies for lead poisoning). - Lay–expert co-production: A local farmer hosting experiments substantially informed research (site-specific plant knowledge, micro-site conditions), accelerating operationalization and shaping research questions. His role exemplified reciprocal knowledge exchange. - Usable science via push–pull: Iterative collaboration with ADEME (e.g., EMETER 2005–2008; APTITUDE 2009–2012; SyMetal 2010–2014) improved usability. Researchers traced dust sources to the dump (contradicting INERIS), which influenced ADEME collaborations and remediation decisions; a startup from the research team later implemented phytostabilisation using local seed sourcing and plant–microbe associations. - Controversy and choice of method: Rival research strands (phytoextraction vs phytostabilisation) and media narratives swayed public opinion and the mayor toward phytoextraction. ADEME selected phytostabilisation, citing robustness, operational feasibility, and timescales; attempts to combine methods failed. - In/visibility in collaborations: Tacit agreements govern evolving goals; strategic and unconscious invisibilisation occurs. Some locals withheld knowledge (e.g., to avoid agromining outcomes); some researchers ceased communication post-fieldwork, generating feelings of dispossession among locals in other projects. - Empowerment and repurposing: Direct interactions and dissemination empowered inhabitants (recognition of indicator plants, counter-expertise on remediation quality). The farmer leveraged seed collection of hyperaccumulators as income and identity reconstruction (vulnerability-as-resilience). Scientific maps of plant collection points aided heritage seekers to locate old mines. - Landscape and memory effects: Phytostabilisation’s greening can invisibilise toxicity and parts of mining heritage, eliciting mixed reactions—some relief (dust control, aesthetic improvement), others’ anxiety and distrust (problem perceived as hidden). Greening supports narratives that may symbolically 'purify' mining legacies and align with national mining renewal discourse. Key data points: Lead up to 30 g/kg (≈600× background) at Les Avinières; 17 scientific and 52 media articles reviewed; research projects: EMETER (2005–2008), APTITUDE (2009–2012), SyMetal (2010–2014).

The findings elucidate how visibilisation/invisibilisation mechanisms shape the entire research–policy–society chain in post-mining contexts. By privileging soil contamination and phytoremediation, research made certain phenomena, stakeholders (e.g., a farmer), and techniques visible while sidelining others (e.g., broader health concerns). Tacit, trust-based collaborations facilitated co-production and usability of science, culminating in ADEME’s adoption of phytostabilisation. Conversely, strategic withholding and perceptions of extractive knowledge practices produced feelings of dispossession, highlighting ethical and procedural challenges. The greening of landscapes via phytostabilisation simultaneously mitigates dispersion risks and obscures ongoing toxicity and historical memory, reinforcing the need for reflexive communication and heritage-sensitive remediation. Overall, the study argues that researchers’ choices actively configure problem framings and governance outcomes, underscoring their social responsibility and the value of reflexive, participatory approaches to manage contested, invisible pollutants.

The paper shows that research in Saint-Laurent-Le Minier both reveals and conceals: it focuses attention on soil contamination, elevates particular lay knowledges and actors, and promotes certain technological pathways (phytostabilisation). Visibilisation and invisibilisation are negotiated through tacit, evolving agreements among stakeholders and are consequential for remediation choices, empowerment, and territorial memory. The authors advocate for a reflexive researcher stance to enhance knowledge transfer, ethical engagement, and social impact. Future work should broaden beyond a microsociological lens to compare multiple post-mining territories and integrate national/EU policy and legal contexts to better understand invisibilisation processes. The recent French Council of State decision regarding nearby Saint-Félix-de-Pallières raises forward-looking questions about technical choices and whether measures will merely invisibilise pollution under operator-led management.

- Sampling and access bias toward the most visible projects and highly engaged inhabitants, leading to rapid data saturation due to the small pool of informed participants. - Sensitivity of topics produced hesitancy to speak on record; reliance on informal interactions to capture key dynamics, which may limit verifiability. - Potential invisibilisation induced by researchers’ own methodological choices (interviewee selection, questions prioritized/omitted). - Language and privacy constraints restrict data availability (interviews in French; GDPR).