Artisanal and small-scale mining (ASM) is a globally significant livelihood activity, supporting millions in low- and middle-income countries. A considerable portion of this mining occurs in biodiversity hotspots, raising concerns about potential environmental damage. While ASM can lead to habitat loss, deforestation, water pollution, and increased pressure on other resources like bushmeat, robust quantification of its biodiversity impacts remains scarce. Most existing evidence relies on descriptive case studies rather than counterfactual analyses. This study focuses on the 2016 sapphire rush at Bemainty in Eastern Madagascar, a high-profile event that garnered significant media attention due to claims of widespread deforestation and threats to lemur populations within the protected Corridor Ankeniheny-Zahamena (CAZ). The study aims to rigorously assess the actual impact of this mining rush on forest loss and degradation using a synthetic control method, complemented by field data from interviews and a lemur survey. This approach is crucial for understanding the nuanced environmental effects of ASM and informing evidence-based policy decisions.

The literature on ASM's environmental impact often presents descriptive accounts from case studies, with quantitative evidence limited and focused mainly on gold mining in regions like the Amazon and Ghana where mercury use is extensive. Studies using satellite imagery have shown varying rates of forest loss in ASM areas, but they lack counterfactual analyses to compare mining impacts to alternative land uses. The Bemainty sapphire rush itself has been subject to conflicting narratives, with some sources claiming hundreds of hectares of deforestation while others suggest pre-existing agricultural clearing as the primary driver of land-cover change. This lack of consensus underscores the need for rigorous evaluation methods.

To evaluate the impact of the Bemainty mining rush, the researchers employed a synthetic control method to estimate counterfactual forest loss in the absence of mining. This involved creating a weighted average of control drainage basins similar to Bemainty in pre-mining characteristics and forest loss patterns. Drainage basins were chosen as the unit of analysis because they influence both gemstone distribution (alluvion deposits) and the potential spread of forest impacts. Three measures of forest loss (raw hectares of deforestation, deforestation rate, and cumulative deforestation) were analyzed at two spatial scales (CAZ and the wider Toamasina province). The analysis included pre-intervention (1991-2011) and post-intervention (2012-2021) data from the Tropical Moist Forests product (TMF), accounting for potential errors using a 1990 forest cover map of Madagascar. Placebo tests were used to assess the significance of any differences between Bemainty and its synthetic control. In addition to the quantitative analysis, the study included informal interviews with 73 residents (miners and farmers) and a lemur survey conducted two years after the mining rush ended. The interviews aimed to gather qualitative information on natural resource use, perceptions of environmental change, and the impact of mining on local livelihoods. The lemur survey involved 27 transects to assess lemur community health and diversity.

The study found no evidence that the Bemainty mining rush significantly increased deforestation or forest degradation compared to the counterfactual scenario. While deforestation increased between 2016 and 2017, this increase was within the range of statistical noise established through placebo tests and consistent with trends in other similar basins. The analysis across different outcome measures and spatial scales consistently showed no significant mining effect. Furthermore, post-rush field data revealed a healthy and diverse lemur community, including populations of critically endangered species. Interviews suggested limited direct deforestation by miners, with most tree harvesting for firewood and construction involving small or dry trees. While some respondents attributed environmental degradation to mining, others (especially miners) emphasized the local community's role in deforestation for shifting agriculture. The per-capita deforestation impact of mining was far smaller than that of agriculture. Although some soil disturbance and water turbidity were observed, the study didn't assess the impact on freshwater biodiversity.

The findings contradict previous claims of extensive deforestation caused by the Bemainty mining rush. The limited impact is attributed to several factors: the geological characteristics of the alluvial sapphire deposits restricted mining activity to a narrow valley area; much of the valley floor was already cleared for agriculture; the mining rush was relatively short-lived; mining activities involved limited tree harvesting and did not require chemical inputs; and the significantly larger footprint of deforestation for agriculture overshadowed any effect from mining. The apparently healthy lemur populations, despite the influx of miners, support the conclusion of minimal ecological trade-offs, potentially due to the persistence of traditional taboos protecting lemurs. While acknowledging limitations such as the retrospective nature of the interviews and potential social desirability bias, the combined quantitative and qualitative evidence strongly indicates that the mining rush's environmental impact on forests was limited.

This study demonstrates that a large-scale ASM rush did not lead to increased forest loss in the Bemainty region, contrary to initial reports. The results highlight the highly variable impact of ASM depending on specific contexts. Future research should focus on high-resolution data to capture smaller-scale impacts, incorporate more comprehensive social and ecological data, and expand the range of ASM activities studied to better inform context-specific policies for balancing socio-economic benefits and environmental protection.

The study's limitations include reliance on self-reported data from interviews (potential for social desirability bias), the relatively small sample size of the interviews, and the fact that the lemur surveys and interviews were conducted two years after the peak of the mining rush. While the synthetic control method helps mitigate some confounding factors, it does not entirely account for all potential unobserved influences on forest loss. Furthermore, the study did not comprehensively assess impacts on freshwater biodiversity.